Olestra is the common name for the mixture of hexa-, hepta- and octaesters of sucrose formed from long-chain fatty acids derived from edible oils. Olestra (Olean, Procter & Gamble, Cincinnati, OH) has taste and cooking characteristics similar to those of traditional fats and oils (Bernhardt 1988, Kester 1993), but does not contribute any energy to the diet because it is not hydrolyzed by gastric lipases and therefore is not absorbed (Mattson and Volpenhein 1972, Miller et al. 1995). Because of these unique properties, olestra can serve as a zero-calorie replacement for conventional fats and oils.

Because olestra is lipophilic and is not absorbed, it can interfere with the absorption of lipophilic nutrients. This interference occurs because the olestra in the gastrointestinal (GI)⁴ tract competes with the intestinal micelles for the lipophilic nutrients. Lipophilic molecules that partition into the olestra are not incorporated into the mixed micelles and transported to the intestinal surface (Jandacek 1982). Consistent with this mechanism, studies in normal healthy human subjects showed that olestra reduced serum concentrations of carotenoids, -tocopherol, and 25-hydroxyergocalciferol [25(OH)D₂ ], the metabolite of dietary vitamin D, but did not affect vitamin K status (Jones et al. 1991a and 1991b, Koonsvitsky et al. 1997). Decreases in serum cholesterol concentration (Crouse and Grundy 1979, Fallat et al. 1976, Glueck et al. 1979) and increases in fecal cholesterol excretion (Jandacek et al. 1980 and 1990) have been seen in humans consuming olestra. Studies in the rat showed that olestra increased fecal cholesterol excretion (Mattson et al. 1976) and reduced liver vitamin A stores (Mattson et al. 1979). Studies in the domestic pig showed that olestra reduced serum and liver concentrations of retinol and -tocopherol, and serum concentration of 25(OH)D₂ (Cooper et al. 1997c, Daher et al. 1997).

The partitioning mechanism would lead one to expect that the effect of olestra on fat-soluble nutrients could be offset by introducing additional amounts of the affected nutrients to olestra or olestra-containing foods. Studies in the pig (Cooper et al. 1997a and 1997b) and in humans (Koonsvitsky et al. 1997, Schlagheck et al. 1997a) have demonstrated this outcome.

The purpose of this study was to determine the dose response of olestra on selected nutrients with a broad range of lipophilicity, in an animal model that closely resembles humans with respect to the handling of these nutrients. The domestic weanling pig was chosen as the animal model. The pig is an appropriate model to use for such a study because of the similarity of the GI anatomy and physiology and requirements for fat-soluble vitamins of pigs and humans; there is also an extensive database on the pig's nutrient requirements and metabolism (Miller and Ullrey 1987). In addition, the pig's vitamin stores and nutritional indices are responsive to dietary changes. The weanling pig has a rapid growth rate and therefore high nutrient demands. Because of this, reductions in nutrient availability are easily detected in the growing pig. Previous studies have shown that the domestic pig is an appropriate model in which to evaluate the nutritional effects of olestra (Cooper et al. 1997c, Daher et al. 1997).

On the basis of the partitioning mechanism (Jandacek 1982), the nutrients with the greatest potential to be affected by olestra are the fat-soluble vitamins. Therefore the effect of olestra on the status of vitamins A, D, E and K was investigated. The partitioning mechanism would lead one to predict that olestra would not affect the absorption of water-soluble nutrients. Nevertheless, the status of selected water-soluble micronutrients was monitored in this study to provide assurance that olestra does not affect these nutrients. Folate and vitamin B₁₂ were chosen as examples of water-soluble nutrients that are digested and absorbed by complex multistep processes; hypothetically, such processes provide the opportunity for olestra to interfere with cleavage and binding reactions. In addition, these nutrients are eaten in microgram amounts, whereas olestra is eaten in gram amounts, thus increasing the possibility that olestra might interfere with their absorption.

Calcium, iron and zinc were chosen as examples of water-soluble micronutrients that many people consume at levels that do not exceed and often do not meet recommended dietary intakes. Therefore any decrease in the bioavailability of these micronutrients would be of potential nutritional concern.

MATERIALS AND METHODS

During the acclimation period, the animals were housed three to five per pen and were observed daily for abnormalities indicative of ill health. They underwent a complete physical examination. Hematology and clinical chemistry measurements were made, and body weights were taken. During the treatment period, the pigs were housed in individual pens in a sunlight-free barn with controlled temperature (above 18°C), ambient humidity, and a 12-h ligh:dark cycle. The pens were cleaned once or twice daily to reduce the potential for coprophagy.

The groups were fed a purified diet (ICN Biomedicals, Cleveland, OH) providing the NRC requirements of micronutrients for 5- to 10-kg swine (National Research Council 1988) and containing 0 (control), 1.1, 2.2, 3.3, 4.4, 5.5 or 7.7% (wt/wt) olestra for 12 wk.

The compositions of the diets are shown in Table A in the Appendix. The purified basal diet fed to the control group furnished 30% of energy from fat and contained 140 g fat/kg of diet. This diet is the same as that fed in other pig studies and has been shown to produce growth equivalent to that produced by standard swine diet (Cooper et al. 1997c, Daher et al. 1997). Vitamins and minerals were added to the diet as a premix, as described in Cooper et al. (1997c).

The dietary concentrations of nutrients and olestra and the homogeneity of the diets were confirmed by analysis. Because of a formulation error, the diets provided only 5% of the requirement for vitamin B₁₂. This did not affect the integrity of the study, for reasons discussed later.

The stability of olestra in the diets was confirmed as described in Cooper et al. (1997c). Stability of dietary nutrients was not determined. Previous analyses showed no instability of nutrients in the same purified diet over a 4-wk period (Cooper et al. 1997c). The diets fed in this study were prepared in sequential batches so that the pigs never received diets that were older than 4 wk.

The olestra was prepared by the method of Rizzi and Taylor (1978) and consisted of 80% octaesters and 20% heptaesters. The relative composition of the fatty acid chains was 20% palmitic, 5% stearic, 38% oleic, 28% linoleic, 7% behenic and 2% others, the same as that used in other pig studies (Cooper et al. 1997a-c, Daher et al. 1997). Before being added to the diet, the olestra was used to fry potato chips under conditions representative of commercial frying processes. This was done to ensure that the olestra fed the pigs was thermally stressed at least to the same degree as the olestra eaten by humans in savory snacks.

The lowest dietary concentration of olestra, 1.1%, was chosen to provide the pigs, at the start of the study, with a daily olestra intake of about twice the estimated mean chronic human intake of olestra from savory snacks, 3.1 g/d (Webb et al. 1997). On the basis of the normal growth of these crossbred pigs (Martin and Crenshaw 1989), the pigs in this group were expected to be consuming about eight times that daily intake by the end of the study. The highest concentration, 7.7%, was judged to be the maximum amount that could be fed without introducing nutritional deficiencies resulting from dilution of the diet (Borzelleca 1992). This concentration would provide the pigs with a daily intake at the start of the study that would be about 18 times the average chronic human intake, or about 8 times the 90th-percentile chronic intake of olestra from savory snacks, 6.9 g/d (Webb et al. 1997). The inclusion of extreme dietary concentrations of olestra increased the opportunity to see effects on water-soluble micronutrients, if such effects existed.

All pigs were exposed to 2 min/d of UV light (FS-40, T12-UVB-U, National Biological, Twinsbury, OH) to mimic the effect of sunlight on vitamin D synthesis. The 2-min daily exposure period was chosen to produce a 50-80% contribution of vitamin D₃ to total vitamin D status (Cooper et al. 1997c). A 50% contribution of vitamin D₃ to total vitamin D status models the worst-case dependence on dietary vitamin D in humans, paralleling the situation for elderly people living in northern latitudes in winter (Delvin et al. 1979).

Table 1. Biological response measured at scheduled intervals in pigs fed up to 7.7% olestra for 12 wk [View Table]

Blood was collected at wk 0, 4, 8 and 12. The samples were taken from the cranial vena cava before the morning feeding. Serum or plasma was stored at 20°C until analyzed. Serum was analyzed for all-trans-retinol (vitamin A), -tocopherol (vitamin E), 25(OH)D₂ , 25-hydroxycholecalciferol [25(OH)D₃ ], and 1,25-dihydroxyvitamin D [1,25(OH)₂D]. In addition, a complete battery of clinical chemistry and hematology parameters was measured, including total iron concentration, total iron-binding capacity (TIBC), total zinc concentration and total calcium concentration. Folate concentration and prothrombin time (PT) were also measured.

The liver was removed from each pig killed at wk 0 (base-line group) or wk 12 through an incision in the cranial abdomen and was perfused with PBS. The entire left lateral lobe was frozen in dry ice and homogenized. Portions of the powder were analyzed for all-trans-retinol, -tocopherol, vitamin B₁₂ (cyanocobalamin), iron and zinc.

Biopsies of adipose tissue were taken from the interscapular region at wk 0, 4, 8 and 12 and analyzed for -tocopherol. The fifth lumbar vertebra was collected from each pig killed at wk 0 or 12, ground into a fine powder and analyzed for total ash content and zinc, calcium and phosphorus concentrations.

The concentrations of vitamin A (total retinol and retinyl esters) in liver and of vitamin E (-tocopherol) in liver and adipose tissues were measured by HPLC following the method of Kayden et al. (1983). Samples were saponified with ethanolic KOH, and the vitamins were extracted with hexane. Quantitation was by HPLC using a silica column (Zorbax, 5-µm, Dupont, Wilmington, DE) and UV detection. Retinol was detected at 325 nm; -tocopherol was detected by fluorescence excitation at 292 nm and emission at 325 nm. USP standards of all-trans-retinol and -tocopherol were used to generate standard curves.

Concentrations of vitamins A and E in serum were measured by HPLC, following the method of Driskell et al. (1982). Serum samples were deproteinized with ethanol and the vitamins extracted with hexane. Aliquots of extracts were reconstituted in hexane and injected onto an HPLC system equipped with a silica gel column (Zorbax, 5-µm, DuPont) and a UV detector. Quantitation was by regression analysis using standard curves generated with USP standards of all-trans-retinol and -tocopherol.

Serum concentrations of 25(OH)D₂ and 25(OH)D₃ were measured by HPLC, following the method of Kao and Heser (1984). The serum samples were acidified with concentrated HCl, and the 25-hydroxy metabolites were extracted on prepacked octadecylsilano silica cartridges (C-18 Bond Elute, Analytichem International, Harbor City, CA). The extracts were purified further by reextracting on aminopropyl cartridges (NH₂ Bond Elute, Analytichem International). The two 25-hydroxy metabolites were quantified simultaneously with a silica column (Zorbax, 5-µm, Dupont) and UV detection (Kratos Analytical, Ramsey, NJ). Concentrations of the two metabolites were calculated from a single calibration curve established with a 25(OH)D₃ standard (Duphar, Amsterdam, The Netherlands). Recovery was determined for each serum sample by adding a 25-hydroxy-(25[27]-methyl-³H)cholecalciferol standard (Amersham, Arlington Heights, IL) to the sample before extraction.

The serum concentration of 1,25(OH)₂D was measured by a radioreceptor method using a thymus receptor specific for both 1,25-dihydroxyergocalciferol [1,25(OH)₂D] and 1,25-dihydroxycholecalciferol [1,25(OH)₃D] (Reinhardt et al. 1984). Briefly, serum samples were spiked with [³H-26,27]-1,25(OH)₂D (INCSTAR, Stillwater, MN), extracted with acetonitrile, and purified further with activated C₁₆OH cartridges (INCSTAR), following the method of Hollis (1986). Bound and free hormones were separated by incubation with charcoal. After incubation, the supernatants containing the bound hormone were decanted into scintillation tubes and the radioactivity content determined with a scintillation counter (Model 1500, Packard Instruments, Downers Grove, IL).

Plasma prothrombin time was measured as part of the clinical chemistry battery.

Plasma concentration of folate was measured by RIA (Matte and Girard 1989, O'Connor et al. 1989). Alkaline denaturation was used to free the folic acid from carrier proteins in plasma. ¹²⁵I-labeled folic acid (Diagnostics Products, Los Angeles, CA) in buffered dithiothreitol was added, and the samples were incubated with -lactoglobulin immobilized on cellulose particles (Diagnostics Products). After incubation, the samples were centrifuged and the radioactivity content of the pellets was determined with a gamma counter (Packard Model 5780). Quantitation was accomplished by the use of a standard curve.

The liver concentration of cyanocobalamin was measured by a microbiologic assay (Baker and Frank 1968, Baker et al. 1986). Aliquots of liver homogenates were diluted 1:10 with 0.25% aconitic acid/1.25% sodium cyanide buffer and autoclaved. After centrifuging to remove debris, the supernatant was diluted again with the aconitic acid/sodium cyanide buffer. Aliquots were added to Erlenmeyer flasks containing the basal growth medium and were sterilized by autoclaving for 20 min. After being cooled to room temperature in a sterile room, each flask was incubated with Ochromonas malhamensis for 5 d under light at 20°C. Growth was measured with a densitometer and quantified by comparison with a standard growth curve. Recovery was determined by spiking liver homogenates with a USP sample of cyanocobalamin. The vitamin B₁₂ assay was conducted by Vitamin Diagnostics (Cliffwood Beach, NJ).

To assay liver for iron and zinc, samples of homogenized liver were digested with concentrated nitric acid and hydrogen peroxide. The resulting solution was filtered and analyzed simultaneously for iron and zinc by atomic emission spectroscopy (ARL 3560, Applied Research Laboratories, Sunland, CA), using an inductively coupled plasma as the excitation source (Dahlquist and Knoll 1978).

To assay for calcium, phosphorus and zinc in bone, fat was extracted from the ground vertebrae with pentane. The defatted sample was weighed into a crucible, charred on a hot plate and ashed in a muffle furnace. The crucible was cooled and reweighed to determine ash content. The result was expressed as a percentage of fat-free dry weight. The resulting ash then was treated with concentrated hydrochloric acid, dried, redissolved in hydrochloric acid, filtered and assayed for calcium, phosphorus and zinc by inductively coupled plasma-atomic emission spectroscopy (Dahlquist and Knoll 1978).

When a measured parameter fell below the detection limit of the analytical method, one half of the detection limit rather than zero was used as the value for that particular measurement (Helsel 1990).

All analyses were conducted at the two-tailed 0.05 significance level, using PC SAS Version 6.04 or SAS Version 6.06 software (SAS Institute, Cary, NC).

RESULTS

Table 2. Daily olestra consumption during wk 1, 6 and 12 by pigs fed up to 7.7% olestra for 12 wk1 [View Table]

Table 3. Cumulative energy consumption, cumulative weight gain and digestible feed efficiency for pigs fed up to 7.7% olestra for 12 wk1 [View Table]

Scattered significant differences among the groups were noted in a number of hematology and clinical chemistry parameters (data not shown). The differences were not dose related and were not consistent over time or between males and females. It was concluded that the differences represented normal biological variability.

Liver concentration of vitamin A (retinol) decreased in a dose-responsive manner with increasing concentrations of dietary olestra (Table 4). The liver vitamin A concentration for the pigs fed 1.1 or 7.7% olestra was about 36 and 7% of control, respectively. Liver vitamin A concentration for the pigs in the control group, 63.3 nmol/g, was about 29% greater than the value measured for the group killed at base line, 49.2 nmol/g.

Table 4. Liver and serum vitamin A concentrations for pigs fed up to 7.7% olestra for 12 wk1 [View Table]

Serum concentration of vitamin A also decreased in a dose-responsive manner with increased dietary olestra concentration (Table 4). The dose response was complete within 8 wk (Table B in the Appendix). At wk 8, the serum concentration for the pigs fed 1.1 or 7.7% olestra was about 63 and 33% of control, respectively. By wk 12, the serum vitamin A concentration increased slightly in all olestra-fed animals, both in absolute terms and in relation to control.

The liver concentration of vitamin E (-tocopherol) decreased with increasing dietary concentration of olestra (Table 5). The mean concentration for the pigs fed 1.1% olestra was about 41% of the control value; for the pigs fed 7.7% olestra, it was about 20% of the control value. The pigs in the control group increased their liver vitamin E stores by a factor of almost three during the study in relation to the value in the animals killed at base line (18.3 vs. 6.69 nmol/g). The pigs fed 1.1% olestra increased their liver vitamin E stores by about 13% over the course of the study.

Table 5. Liver, serum, and adipose vitamin E concentrations for pigs fed up to 7.7% olestra for 12 wk1 [View Table]

Serum vitamin E concentration decreased in a dose-responsive manner with increasing dietary concentration of olestra, paralleling the change in the liver concentration of the vitamin (Table 5). The dose response was essentially complete within 4 wk (Table C in the Appendix). The largest effect was measured at wk 8, when serum vitamin E concentration for the pigs fed 1.1 or 7.7% olestra was about 43 and 18% of control, respectively. By wk 12, serum vitamin E concentration was increasing in all olestra-fed groups, both in absolute values and in relation to control.

Serum vitamin E concentration normalized with respect to serum lipids showed the same response to olestra as did the nonnormalized concentration (data not shown).

The concentration of vitamin E in adipose tissue decreased in a dose-responsive manner with increasing dietary concentration of olestra (Table 5). Unlike serum vitamin E, adipose vitamin E continued to decrease between wk 8 and 12 for the olestra-fed groups (Table D in the Appendix). At wk 12, the adipose vitamin E concentration for the pigs fed 1.1 or 7.7% olestra was about 52 and 27% of control, respectively.

At wk 12, serum 25(OH)D₂ decreased in a dose-responsive manner with increasing olestra dietary concentrations up to 3.3% but tended to increase with further increases in olestra dietary concentrations (Table 6). The dose response was present at olestra dietary concentrations <5.5% at wk 4 and <4.4% at wk 8 (Table E in the Appendix).

Table 6. Serum 25-hydroxyergocalciferol [25(OH)D2], serum 25-hydroxycholecalciferol [25(OH)D3], total 25-hydroxyvitamin D [25(OH)D] and 1,25-dihydroxyvitamin D [1,25(OH)2D] concentrations for pigs fed up to 7.7% olestra for 12 wk1 [View Table]

At wk 12, serum 25(OH)D₃ concentration for the group fed 1.1% olestra was significantly greater than the control value (Table 6). The group fed 7.7% olestra had a mean serum 25(OH)D₃ concentration which was significantly less than the control value (Table 6). Similar trends were observed at wk 4 and 8 (Table F in the Appendix).

Serum concentrations of total 25(OH)D reflected changes in the serum concentration of 25(OH)D₃ , which contributed from 87 to 89% of the 25(OH)D concentration (Table 6 and Table G in the Appendix). No olestra dose response was found on the serum concentration of 1,25(OH)₂D (Table 6). A time-dependent decrease in serum 1,25(OH)₂D was noted in all groups, including the control group (Table H in the Appendix).

No effect of olestra was noted on PT (Table 7). PT measurements were made weekly for the control, 5.5% olestra and 7.7% olestra groups. No significant differences were found among the PT values at any of these time points (data not shown).

Table 7. Plasma prothrombin time for pigs fed up to 7.7% olestra for 12 wk1 [View Table]

No olestra dose response on plasma folate concentration was found (Table 8). At wk 4 and 8, plasma folate concentrations for some olestra groups were significantly different than control: some were less, some greater.

Table 8. Plasma folate concentration for pigs fed up to 7.7% olestra for 12 wk1 [View Table]

Liver vitamin B₁₂ concentration for the 7.7% olestra group was significantly less than the control value and thus produced a significant trend (Table 9). However, mean corpuscular volume (MCV), which normally is elevated when vitamin B₁₂ is deficient (Herbert 1988), was 54 ± 2.5 fL for both males and females in the 7.7% olestra group, and 52 ± 3.5 and 55 ± 2.0 fL for males and females, respectively, in the control group (complete MCV data are not shown). The liver concentration of vitamin B₁₂ measured at wk 12 in the control group, 64.2 ± 13.3 nmol/g liver, was not significantly different than the concentration measured in the group killed at base line, 80.4 ± 16.2 nmol/g liver (Table 9).

Table 9. Liver vitamin B12 concentration for pigs fed up to 7.7% for 12 wk1 [View Table]

Liver iron concentration for the groups fed 5.5 or 7.7% olestra was significantly less than the control value, resulting in a significant trend test (Table 10). No olestra dose response was found on serum TIBC or total iron (Table 10 and Table I in the Appendix). No olestra effects were noted on hematology parameters such as RBC, mean corpuscular hemoglobin (MCH) and mean corpuscular hemoglobin concentration (MCHC) (data not shown).

Table 10. Liver iron, serum total iron-binding capacity (TIBC) and serum total iron concentration for pigs fed up to 7.7% olestra for 12 wk1 [View Table]

There were no significant differences in liver, bone or serum concentrations of zinc between the olestra-fed and control groups and no significant trends in the data (Table 11 and Table J in the Appendix). A significant downward trend in bone ash was found with increasing dietary olestra concentration (Table 12). The value for the group fed 4.4% olestra was significantly less than the control value because one pig had a bone ash value of 55.3%. The mean for the rest of the animals in the group was 60.5 ± 1.2%, not significantly different than the control value of 61.1 ± 1.0%. No effect of olestra on bone concentrations of calcium or phosphorus occurred (Table 12).

Table 11. Liver, bone and serum zinc concentrations for pigs fed up to 7.7% olestra for 12 wk1 [View Table]

Table 12. Bone ash content, bone calcium and phosphorus concentration in bone for pigs fed up to 7.7% olestra for 12 wk1 [View Table]

Olestra had no significant effects on serum concentration of calcium or inorganic phosphorus (Table 13 and Table K in the Appendix). Alkaline phosphatase was also unaffected by olestra intake (data not shown).

Table 13. Serum calcium and inorganic phosphorus concentrations for pigs fed up to 7.7% olestra for 12 wk1 [View Table]

DISCUSSION

Olestra was well tolerated by the pigs as evidenced by the lack of any health-related antemortem observations or changes in clinical chemistry or hematology measures. The rate of growth was the same for all groups and was essentially the same as that observed when pigs are fed a standard corn-soy-based swine diet (Martin and Crenshaw 1989).

The absorption or utilization of macronutrients was unaffected by olestra as evidenced by growth and the lack of any effect on the amount of digestible energy consumed. These findings are in agreement with those from other long-term animal feeding studies conducted as part of the safety evaluation of olestra (Lafranconi et al. 1994, Miller, K. et al. 1991, Wood et al.1991).

In agreement with findings from other studies, there was no effect of olestra on absorption of water-soluble nutrients despite the exaggerated conditions of the study (Cooper et al. 1997a, Schlagheck et al. 1997a and 1997b). The lack of an effect of olestra on the absorption of water-soluble nutrients is consistent with the hypothesis that the only way olestra interferes with nutrient absorption is via the partitioning mechanism, i.e., lipophilic nutrients partition into the olestra and become unavailable to the intestinal micelles (Jandacek 1982).

Olestra affected the absorption of vitamins A, D₂ and E as expected on the basis of the partitioning mechanism and as observed by others (Glueck et al. 1982, Jones et al. 1991b, Koonsvitsky et al. 1997, Mattson et al. 1979, Mellies et al. 1983 and 1985, Schlagheck et al. 1997b). Olestra did not affect vitamin K status in agreement with results from other studies (Cooper et al. 1997b, Jones et al. 1991a, Koonsvitsky et al. 1997, Schlagheck et al. 1997b).

The dose response of olestra on vitamin A was established with dietary sources of the vitamin, i.e., retinyl palmitate and -carotene, similar to the sources in the U.S. diet (Olson 1988). The effect on the liver concentration of vitamin A can be interpreted as an effect on the absorption of the vitamin because olestra has been shown not to affect the irreversible loss of vitamin A from normal metabolism (Food Additive Petition 7A 3997). In addition, the expansion of the liver pool size was the same in all groups, another factor that could affect the liver concentration of the vitamin.

The expansion in the size of the liver pool during growth of the pigs in this study had a substantial effect on the liver concentration of vitamin A. However, the expansion did not affect the dose response because it occurred to essentially the same degree in all groups. Calculations shown that the pigs killed at base line had about 11,760 nmol (12,000 g BW × 2 g liver/100 g BW × 49 nmol vitamin A/g liver) of vitamin A liver stores, assuming that liver represents about 2% of body weight (Filer et al. 1966). At the end of the study, the pigs fed 1.1% olestra, the dietary concentration most closely related to human intake, had vitamin A liver stores of about 33,580 nmol. Although the concentration of vitamin A in the liver was about 53% less for this group than for the group killed at base line, there was almost a threefold net gain in total vitamin A stores.

Serum vitamin A concentration decreased slightly for the control group during the study, from 0.86 to 0.64 µmol/L. Serum vitamin A concentration has been found to drop in pigs after weaning (Miller, E. et al. 1991). The dose response of olestra on serum vitamin A observed here was not surprising in view of the liver concentration of the vitamin. Serum vitamin A reflects differences in intake of the vitamin when liver vitamin A is less than ~70 nmol/g liver, as it was in this study (Hentges et al. 1952, Olson 1984, Sauberlich et al. 1974).

In humans, functional impairment, including impaired dark adaptation, night blindness and ocular lesions, can occur when serum vitamin A falls below ~0.35 µmol/L (10 µg/dL). For example, corneal xerophthalmia occurs commonly in Indian and Indonesia children with serum vitamin A concentrations < 0.35 µmol/L (Gibson 1990). However, in pigs, visible signs of vitamin A deficiency do not appear unless serum vitamin A concentrations are <0.18 µmol/L. Hentges et al. (1952) observed incoordination of movement, paresis of hind legs, tilted heads and night blindness in pigs only after plasma vitamin A concentrations fell below 0.18 µmol/L. In this study, the lowest average serum vitamin A concentration measured among the pigs fed olestra was 0.23 ± 0.07 µmol/L (Table 4). None of the pigs showed clinical signs of vitamin A deficiency and there was no decline in growth or feed intake among the olestra-fed groups.

Liver, serum, and adipose concentrations of vitamin E responded similarly to olestra intake, consistent with a reduction in absorption efficiency. Each measure was affected essentially in the same way, although adipose concentration of the vitamin responded more slowly than liver or serum concentrations. Because olestra affects only the absorption of vitamin E, all pools of the vitamin would be expected to respond similarly in growing animals. Bieri (1972) found that plasma, liver and other tissue concentrations of vitamin E declined rapidly in weanling rats fed vitamin E-deficient diets; most of the change occurred within 2 wk. Machlin et al. (1979) found that changes in availability of dietary vitamin E produced parallel changes in plasma and tissue concentrations within a few weeks in young guinea pigs.

The concentration of vitamin E in adipose tissue changed more slowly than serum and liver concentrations in response to olestra because of a slower expansion of this pool, in relation to lean tissues, in pigs in this weight range (Shields et al. 1983). Machlin et al. (1979) found that the adipose pool of vitamin E decreased more slowly than other tissue pools in young guinea pigs fed a vitamin E-free diet.

At olestra dietary levels below 3.3-5.5%, depending when the measurement was made (wk 4, 8 or 12), the responses of serum 25(OH)D₂ and 25(OH)D₃ to olestra were as expected, i.e., serum 25(OH)D₂ declined in a dose-responsive manner and serum 25(OH)D₃ was unaffected. The shape of the 25(OH)D₂ response was similar to the responses observed for vitamins A and E, indicating that olestra interacts with vitamin D in the same way as it interacts with the other fat-soluble vitamins.

The increase in serum 25(OH)D₂ and the decrease in serum 25(OH)D₃ observed with long-term consumption of high dietary levels of olestra were unexpected. Such effects were not seen in pigs fed up to 5.5% olestra for 26 wk without exposure to UV light (Cooper et al. 1997b) or in human subjects consuming up to 32 g/d olestra (Schlagheck et al. 1997b). A plausible explanation for the reciprocal changes in serum concentrations of the two metabolites is competition between vitamin D₃ and vitamin D₂ for liver 25--hydroxylase. In pigs, vitamin D₃ is the preferred substrate for 25--hydroxylase (Horst et al. 1982); therefore an increase in serum 25(OH)D₂ might occur if the amount of vitamin D₃ available for hydroxylation decreased.

The declines in serum 25(OH)D₃ and serum 25(OH)D noted with long-term consumption of 5.5 and 7.7% olestra may have resulted from an interference with resorption of the enterohaptic-circulating forms of vitamin D₃. In these pigs, almost 90% of the serum 25(OH)D concentration came from 25(OH)D₃ . Vitamin D₃ and 25(OH)D₃ are secreted in the bile, and some fraction is reabsorbed from the intestine (Arnaud et al. 1975, Avioli et al. 1967, Nagubandi et al. 1980). A similar change in 25(OH)D₃ has been reported for humans eating large amounts of dietary fiber (Batchelor and Compston 1983), and reductions in vitamin D status have been observed in clinical malabsorption syndromes in which the resorption of biliary-derived substances is hindered (Batchelor et al. 1982, Compston et al. 1982).

The decline in serum 1,25(OH)₂D concentration noted in all groups was a result of aging (Horst and Littledike 1982, Lachenmaier-Currle and Harmeyer 1988). Serum 1,25(OH)₂D is tightly regulated by calcium, phosphorus and parathyroid hormone (PTH) and is a reliable indicator of changes in dietary vitamin D intake only under deficiency conditions (Schrijver 1991).

The lack of an effect of olestra on PT agrees with results from other pig (Cooper et al. 1997a and 1997b) and human studies (Jones et al. 1991a, Koonsvitsky et al. 1997, Schlagheck et al. 1997a and 1997b). Any effect of olestra on the absorption of phylloquinone, and measurements of serum phylloquinone concentrations in humans indicate that there is such an effect (Schlagheck et al. 1997a and 1997b), is not sufficient to affect vitamin K functional status.

Although vitamin B₁₂ was fed at only 5% of the NRC requirement, the liver concentrations were in the range known to be responsive to changes in intake of the vitamin (Catron et al. 1952). The amount of vitamin B₁₂ fed was sufficient to allow the pigs in the control group to increase their total pool of vitamin B₁₂ by a factor of almost 5, calculated as discussed above for vitamin A.

The significantly lower liver vitamin B₁₂ concentration for the pigs fed 7.7% olestra (and the resulting significant trend test noted in this study) was probably not an effect of olestra on vitamin B₁₂ absorption. Vitamin B₁₂ absorption was not affected when adult human subjects ate up to 32 g of olestra in a meal along with vitamin B₁₂ (Schlagheck et al. 1997a and 1997b) or when pigs were fed up to 7.7% olestra with graded levels of vitamins A, D, and E for 12 wk, or up to 5.5% olestra with graded levels of vitamins A and E for 26 wk (Cooper et al. 1997a and 1997b). In addition, MCV, which is normally elevated when vitamin B₁₂ is deficient (Herbert 1988), was unaffected in this study, further supporting the conclusion that vitamin B₁₂ absorption was not directly affected by olestra. MCV was 54 ± 2.5 fL for both males and females in the 7.7% olestra group, in which the liver vitamin B₁₂ concentration was the lowest, compared with control values of 52 ± 3.5 fL (males) and 55 ± 2.0 fL (females).

The low liver vitamin B₁₂ concentration for the groups of pigs fed 7.7% olestra was possibly a result of the poor vitamin E status of those pigs. The mean liver vitamin E concentration for that group, 3.53 nmol/g, was only slightly greater than the value of 3.0 nmol/g associated with signs of vitamin E deficiency in pigs (Jensen et al. 1988). It is known that -tocopherol is needed for the conversion of methylcobalamin to adenosylcobalamin (Turley and Brewster 1993), the major form of vitamin B₁₂ stored in the liver (Herbert and Das 1994). The microbiologic assay used to determine the concentration of vitamin B₁₂ in the liver in this study measures the metabolically active forms of the vitamin, including adenosylcobalamin (Baker and Frank 1986, Baker et al. 1986). Pappu et al. (1978) showed that the synthesis of adenosylcobalamin is inhibited in the rat when the liver concentration of vitamin E is low. The fact that no effects on liver vitamin B₁₂ were observed in the pig studies in which extra amounts of vitamin E were added to the diet supports the conclusion that the effects seen in this study were probably related to the vitamin E status of the pigs and not to a direct effect of olestra on vitamin B₁₂ absorption.

The lack of any effect on the status of either vitamin B₁₂ or folate, both water-soluble nutrients absorbed by complex multistep processes (Herbert and Colman 1988), provides assurance that olestra does not interfere with nutrient availability by means other than the partitioning of fat-soluble nutrients into the olestra.

Although the liver concentrations of iron for the pigs fed 5.5 or 7.7% olestra were significantly less than the control value, no effects on serum TIBC or serum total iron concentration or on the hematology parameters RBC, MCH and MCHC were observed for these groups, indicating that the effects on the liver concentration were probably not direct effects of olestra on iron absorption. Other studies in pigs (Cooper et al. 1997a and 1997b) and humans (Schlagheck et al. 1997a and 1997b) showed no effect of olestra on iron absorption. Also, there was no effect of olestra on any of the measures of zinc status of the pigs. If the effect on liver iron concentration for the 5.5 and 7.7% olestra groups resulted from a direct effect of olestra on iron absorption, a similar effect might have been expected on zinc absorption.

The low liver concentration of iron for the pigs fed the highest levels of olestra may have resulted from the poor vitamin A status of those pigs. Vitamin A is required for normal hematopoiesis, and improvements in iron status with vitamin A treatment have been observed among children with marginal iron status (Bloem et al. 1989, Mejia and Chew 1988).

The significant decreasing trend in bone ash with increasing dietary levels of olestra was most likely a secondary effect of the low vitamin A status of the animals. In vitamin A deficiency, deposition of calcium in bone is decreased; this decrease can lead to a decrease in bone ash (Navia and Harris 1980). Other indices of calcium status such as bone and serum calcium and phosphorus concentrations were unaffected, and no effect of olestra has been found in other pig studies. Importantly, when pigs were fed up to 7.7% olestra and graded levels of vitamins A, D and E, no effect of olestra on bone ash content or other measures of calcium status was seen (Cooper et al. 1997a). Similarly, when pigs were fed up to 5.5% olestra for 26 wk and PTH concentration was measured, in addition to the measures used in this study, there was no effect of olestra on any of the calcium status measures (Cooper et al. 1997b). PTH, in addition to 1,25(OH)₂D, is a primary factor in calcium homeostasis (Avioli 1988), and changes in serum calcium concentration lead to reciprocal changes in serum PTH concentration (Silver 1992).

Findings from this study generally agreed with those from other pig and human studies and confirmed expectations based on the partitioning mechanism. Because of the similarities between the GI tracts of weanling pigs and young children (Leary and Lecce 1976), these findings are especially important in understanding potential olestra effects in children. Because of the dietary conditions used in the study, it is unlikely that effects not seen in the pig will be seen in humans. Further, the effects of olestra on the absorption of fat-soluble vitamins will be considerably less in people eating olestra in real-life conditions than those measured in pigs in this study.

ACKNOWLEDGMENTS