Safety Evaluation of Olestra - ACS Publications - American Chemical

Chapter 32

Safety Evaluation of Olestra

Downloaded via UNIV OF CALIFORNIA SANTA BARBARA on July 11, 2018 at 13:23:02 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

A Nonabsorbable Fat Replacement Derived from Fat Carolyn M. Bergholz The Procter & Gamble Company, Cincinnati, O H 45224

Olestra is the mixture of the hexa-, hepta-, and octa-esters of sucrose with long-chain fatty acids from any edible oil. Its physical properties are comparable to those of triglycerides, but it is not digested by lipolytic enzymes or absorbed and therefore is noncaloric. Technically it can replace fat in a wide variety of foods and can be used to make cooked, baked and fried foods lower in fat and calories. A Food Additive Petition is under review by the F D A which is comprised of results of extensive testing in animals and humans. The major areas of investigation are metabolism and absorption, chronic toxicity, mutagenicity, carcinogenicity, reproductive and developmental toxicity, safety for gastrointestinal tract, nutrition, and the potential for olestra to affect absorption of drugs. This testing involved studies in five different species of animals and over 30 clinical investigations. The results of this research support the safety of olestra for use in foods.

Olestra is the common and usual name proposed for the mixture of the hexa-, hepta-, and octa-esters of sucrose formed with long-chain fatty acids from any edible oil. Because its physical properties are like those of a triglyceride, it functions the same as fat in foods and performs the same as fat during cooking or frying. However, olestra is not digested by pancreatic enzymes and therefore is not absorbed from the gastrointestinal tract and contributes no calories. A Food Additive Petition is under review by the Food and Drug Administration (FDA) which requests approval to use olestra in place of fat for preparation of specific foods. This petition is comprised largely of reports of studies on the safety of olestra. The safety evaluation of olestra, like that of any material, is determined by three fundamental considerations: the chemistry of the material, its biological properties NOTE: Please address correspondence to K. D . Lawson, The Procter & Gamble Company, 6071 Center Hill Road, Cincinnati, OH 45224 0097-6156/92/0484-0391$06.00/0 © 1992 American Chemical Society

Finley et al.; Food Safety Assessment ACS Symposium Series; American Chemical Society: Washington, DC, 1992.

392

FOOD SAFETY ASSESSMENT

and the expected exposure. This chapter reviews the olestra safety research program in the context of these considerations. The program is very extensive and is comprised of over 100 laboratory and clinical investigations. It is impossible, obviously, to present the results of the testing that support the safety of olestra in this review chapter. Rather, the intent is to provide an overview of the scope of the program, discuss the major areas of investigation, the most recent data and important conclusions, citing references for studies that have been published. Chemistry The raw materials used to make olestra are sucrose from sugar cane or beet sugar and fatty acids from edible oils. The structure of olestra is analogous to that of a fat, i.e., triglyceride molecule. Instead of a glycerol core with three fatty acid esters, olestra has a larger core, a sucrose molecule, with 6-8 fatty acid esters. The physical properties of olestra, like those of a triglyceride, are determined by the fatty acid composition, which varies depending upon the source oil used. A high proportion of long-chain and/or saturated fatty acids, for example, will increase the viscosity and raise the melting point of the olestra. The safety evaluation of olestra must provide assurance of safety for the full range of compositions specified for food use. Olestra is comprised largely of octaesters, as shown in the specifications listed in Table 1. There are also specifications for minor impurities arising from the starting materials or as by-products of the synthesis. These are the same as those for conventional fats and oils or other fatty acid derived food additives. Table 1. Olestra Ester Distribution Specifications Total Octa-, Hepta-, Hexaester Octaester Hexaester Penta- and Lower esters

> 97% > 70% < 1% < 0.5%

Because the anticipated uses of olestra include frying applications, there has been extensive investigation of the chemistry of heated olestra compared to heated fat. The results show that the olestra sucrose backbone is extremely heat stable and that reactions involving the fatty acid side chains are the same as those that occur during heating of triglyceride. This has been demonstrated by heating olestra for 7 days under deep fat frying conditions and using gas chromatography and mass spectroscopy to compare the fatty acid reaction products with those from the corresponding triglycer ide heated under the same conditions. No unique products were detected at a level of sensitivity of 5 ppm (Henry, D. E.; Tallmadge, D. H.; Saunders, R. Α.; and Gardner, D. R., Procter & Gamble, unpublished data). Biological Properties The most important biological property of olestra is its lack of absorption. This property, of course, is directly related to its chemical structure. Mattson and


32.

BERGHOLZ


393

Volpenhein in 1972 reported that a polyol backbone fully esterified with six or more fatty acids was completely resistant to digestion by pancreatic enzymes (7). One of the polyols tested was sucrose with eight fatty acid esters. Subsequent experiments in animals showed that a polyol with 6-8 fatty acid esters was not absorbed (2). These data and work of others established that fat must first be hydrolyzed to free fatty acids and 2-monoglycerides before it can form mixed micelles with bile salts and be absorbed from the intestinal lumen. The larger bulk of highly esterified sucrose apparently hinders enzymatic cleavage of the fatty acid ester bond in olestra. Nonabsorption has a number of implications for design of a program to evaluate the safety of olestra for food use. First of all, it is important to determine the extent to which no absorption can be established. This is a challenge since it is impossible to prove zero. The kinds of approaches used will be discussed further below. The safety program must of course demonstrate the absence of any systemic toxicity after long-term ingestion. But importantly, nonabsorption implies that the gastrointestinal (GI) tract is the only organ system that is exposed to olestra. Therefore the safety evaluation must thoroughly assess the potential to affect the structure and function of the gastrointestinal system. A third implication of nonabsorption is the existence of a nonabsorbable lipid phase in the GI tract. This raises important questions about the potential to affect absorption of lipophilic constituents of the diet and lipophilic medications. Absorption and Toxicology Understanding the absorption, distribution, metabolism and excretion (ADME) of olestra, as with any new material, is fundamental to any safety evaluation. A number of different approaches have been utilized, all of which provide evidence that olestra is not digested and not absorbed from the gastrointestinal tract. Results of in vitro experiments showed that olestra was not hydrolyzed regardless of the fatty acid chain length or degree of unsaturation (7). Results of material balance studies (2-4) and early experiments with radiolabeled olestra were consistent with nonabsorption. To learn where olestra would go if it were absorbed and if it would be metabolized, it was necessary to use intravenous (IV) dosing (5-7). Results showed that after IV administration of olestra to rats and monkeys, almost 70% of the olestra was rapidly taken up by the liver. Over time, unmetabolized olestra was slowly excreted via the bile into the stool. Therefore, if trace amounts of olestra were absorbed from the gastrointestinal tract, it would most likely accumulate in the liver. Based on this understanding, liver tissue from rats fed olestra at 9% (w/w) of the diet for two years was analyzed for olestra (7). The results showed that olestra did not accumulate in tissues. Based on the sensitivity of the analytical method and the kinetics of excretion, it was calculated that olestra would have been detected in the liver if more than 1 χ 10"6 % of the total amount ingested were present. Similarly, no olestra was detected in any tissues from monkeys after 29 consecutive months of diet with 8% (w/w) olestra. The calculated limit of detection in this study was 4 χ 10"^ % of the dose. Results of stateof-the-art A D M E studies using high specific activity radiolabeled olestra continue to


394


confirm that olestra does not accumulate in tissues above the limit of detection (Miller, K . W., Procter & Gamble, personal communications). The potential for systemic toxicity was evaluated in subchronic and chronic feeding studies conducted in five different animal species with olestra at concentrations up to 15% (w/w) of the diet (Table 2). Results of all of these studies confirm that olestra is not toxic. Body weight gain, urine and blood chemistries, hematology and microscopic examination of all tissues showed no adverse effects related to ingestion of olestra (8-11). Results of two 2-year studies in rats also demonstrated that olestra is not carcinogenic (8). A long-term study in mice is in progress and is expected to again confirm that olestra does not cause tumors. Among these studies was a 91-day study of olestra heated under foodservice deep fat frying conditions (77). Table 2. Feeding Studies Conducted in Animals to Evaluate the Safety of Olestra Species Subchronic: Rat

Hamster Dog Mouse Long-Term: Dog Monkey Rat Mouse

Duration (days)

% (w/w) in Diet

28 91 91 91 (unheated) 28 28 91 91

4,8,15 4, 8,15 4, 8,15 3.5,7.5 5,10 5,10 5,10 2.5,5,10

20 44 24 24 24 (in progress)

5,10 8 1,5,9 9 2.5, 5,10

A series of short-term genotoxicity tests also demonstrated that olestra does not cause mutations, chromosomal aberrations, or affect D N A repair (72). A multi-generation feeding study in rats confirmed that olestra does not affect reproduction or embryonic development (13). In summary, the lack of absorption and accumulation of olestra in tissues has been established at very low limits of detection. The results of a battery of basic toxicity evaluations demonstrate that olestra is not toxic, carcinogenic, or genotoxic and is not a reproductive toxin. It is of interest to note that this toxicity evaluation is atypical in some respects. First, a toxic effect level was not identified. Actually this is not surprising since a nonabsorbed material is far less likely to cause toxicity. Secondly, it is impossible to feed olestra to animals at levels 100-times the intake expected from the use of olestra to replace fat in foods. For example, if expected intake over time is 6 grams per day,


32. BERGHOLZ


395

a rat would have to ingest 600 grams of olestra per day in addition to sufficient diet to meet energy and nutritional needs. For perspective, rats consume 150-200 grams of chow per week. In the feeding studies, olestra was tested at doses which exceeded the maximum high dose recommended for toxicity testing (5% by weight of the diet). The power of animal feeding studies for testing a macronutrient substitute rests on the wide number of parameters that can be measured and the ability to thoroughly examine tissues grossly and microscopically. The absence of adverse effects in any animal species tested in the basic toxicity evaluation opened the door to clinical evaluation of potential adverse health and nutritional effects in humans. Gastrointestinal Safety As discussed above, the GI tract is the only organ system exposed to measurable amounts of olestra. Therefore, the potential to affect the GI system was evaluated in both animal and human studies. In long-term feeding studies in animals, detailed microscopic examination of all segments of the GI tract showed that high dietary concentrations of olestra did not cause morphologic changes (8-11). Special stains and chemical analyses also showed that olestra was not present in associated lymphoid tissues. Clinical studies have shown that the presence of olestra in the GI tract has no effect on gastric emptying time (14), GI motility or transit time through segments of the bowel (Aggarwal, Α., Mayo Clinic, personal communications). A study in rats demonstrated that olestra does not stimulate secretion of pancreatic enzymes (75). Regulation of these GI functions is determined by the presence of free fatty acids from hydrolysis of fat, by calorie density and by water-soluble mediators. Olestra is not hydrolyzed, non-caloric and does not participate in the aqueous or mixed micellar phases in the intestinal lumen. It does not appear to provide signals that regulate digestion or to interfere with signals provided by other components of the diet. Bile acid physiology is another important area of olestra safety research, since bile acids are important to digestion and absorption of fatty acids. Studies in rats, monkeys, and humans showed that excretion of fecal bile acids is normal and that olestra does not alter the composition of bile or reabsorption of bile acids from the intestinal lumen (16,17, St. Clair, R. W., Wake Forest University, unpublished data). Table 3 shows the concentration of primary and secondary bile acids in bile of monkeys fed olestra at 6% of the diet for one year compared to the profile of chow-fed controls. There were no significant differences, indicating that olestra did not affect synthesis or reabsorption of bile acids. Extensive clinical research has shown that ingestion of foods made with olestra does not adversely affect bowel function, even with high intakes of 30-50 grams of olestra per day (4,17-21). Fecal water and electrolyte content is not affected in human or animals, confirming that olestra does not cause diarrhea. Stool consistency is sometimes softer due to the unabsorbed lipid-like material, but this is reported as a benefit (4). One reason bowel function and stool consistency are normal is that olestra has no osmotic effect in the colon and is not metabolized by colonic microflora. Microbial metabolism was investigated by in vitro anaerobic fecal culture systems


396


with radiolabeled olestra. Results showed no degradation of fatty acids or change in ester distribution (22). Material balance studies also showed that olestra is excreted unchanged in feces, indicating that it is not metabolized by gut microflora (4). Table 3. Biliary Bile Acid Composition in Monkeys Fed Olestra Olestra in Diet (% w/w) 6 0

η

Day of Study

Lithocholic Acid

Bile Acid (mol % ±SEM) Chenodeoxycholic Deoxycholic Cholic Acid Acid Acid

5 5 4 3

316 373 316 373

6.6 ± 1 . 9 5.6 ± 1 . 2 4.4 ± . 8 4 3.9 ± . 6 6

52.4 ± 7 . 5 55.1 ± 7 . 5 58.6 ± 7 . 4 56.0 ± 4 . 7

15.6 ± 2 . 7 15.3 ± 3 . 2 13.0 ± 4 . 4 16.5 ± 5 . 7

25.4±7.7 24.0 ±6.1 24.0 ±5.7 23.6 ± 2 . 2

Environmental safety studies have demonstrated that once olestra leaves the body it does not affect water treatment systems. It absorbs to sludge and can then be completely degraded aerobically by soil microorganisms following agricultural application of sludge (Greff, J. Α., Procter & Gamble, personal communications). Nutrition Research The major function of the gastrointestinal system, of course, is digestion and absorption of nutrients. The potential for olestra to affect utilization of nutrients has been of primary importance in the evaluation of the safety of olestra. Utilization of water-soluble macro- and micro-nutrients, i.e., carbohydrates, amino acids, minerals and water-soluble vitamins would not be expected to be affected by olestra. Results of animal and clinical studies are consistent with this assessment. Data from animal studies show that growth and development are normal (8,9,17,73). Fasting blood sugar was not affected by olestra in animal studies (Jandacek, R. J.; Holcombe, Β. N . , Procter & Gamble, unpublished data) or in diabetic patients (27), confirming that olestra does not affect carbohydrate absorption. Other evidence that proteins and minerals are utilized normally in the presence of olestra comes from studies in rats of absorption of radiolabeled hydrophobic amino acids, (Gibson, W. B., Procter & Gamble, unpublished data), clinical studies showing no effect on fecal minerals or fecal nitrogen, and mineral levels in serum chemistries from animal and human studies (811,18-21). Animal research showed very early that absorption of fat-soluble constituents in the diet, in particular cholesterol and fat-soluble vitamins, could be reduced by partitioning into olestra (23£4). The potential for this to occur and the extent to which absorption is decreased depends on a number of variables. Among them are the lipid solubility of the constituent, the solute concentration, the kinetics of water/oil partitioning, absorption kinetics, and importantly, the amount of olestra (16,18-2025). Using radiolabeled cholesterol in olestra, it was demonstrated that 14g/day of olestra


32. BERGHOLZ


397

decreased absorption of dietary cholesterol about 9% (26). Olestra also decreases reabsorption of cholesterol in enterohepatic circulation (76,25). The fat-soluble vitamins vary in lipophilicity. To address the potential for the proposed uses of olestra to affect the nutritional status of fat-soluble vitamins, large base size, double blind, placebo-controlled clinical investigations have been con ducted. The results of a 16-week study showed only a modest reduction in serum vitamin Ε levels after two weeks of ingestion of 18 grams of olestra per day (27). Continued ingestion for a total of 16 weeks gave no further reduction. Importantly, even with the modest reduction, serum alpha-tocopherol was at normal levels throughout the study. In this same 16-week clinical investigation, functional prothrombin was measured by assay of the Simplastin:Ecarin (S/E) ratio. This assay has been shown to detect changes in vitamin Κ intake within days (28). The results of the 16-week investigation showed no change in the S/E ratio throughout the study, indicating 18 g/day of olestra does not affect vitamin Κ status as measured by this assay. These data are consistent with the results of another double-blind placebo-controlled study among 200 subjects which showed no difference from baseline in any measure of vitamin Κ status over six weeks of ingestion of olestra at 20 g/day (29). During the 16-week study, there also was no difference from control in 25-hydroxy vitamin D status. This was predicted based on results of other clinical investigations (18 JO). Vitamin A , as measured by plasma retinol levels, also showed no change, as expected (18). However, plasma retinol levels are not a sensitive measure of potential effects on vitamin A status, since most individuals have large stores of vitamin A in the liver which serve to maintain constant blood levels. Other approaches are being used to determine the dose-response effect on liver stores of vitamin A and potential for food uses to meaningfully affect status. The results of animal and clinical test data must be considered in light of the proposed food uses and levels of fat replacement. These considerations determine the expected amount and pattern of chronic and single day intakes for the general population, various age groups and special subgroups, and are essential for assessment of the appropriateness of supplementation of a particular olestra food with a particular vitamin. The goal is to ensure that foods made with olestra have full nutritional value, with the exception of less fat and fewer calories. Importantly, foods made with olestra will have the same taste and mouthfeel as those made with full calorie fats. Drug Absorption The safety evaluation of olestra has also included studies which established that olestra will not affect absorption and efficacy of lipophilic drugs. In one study (31), a single dose of propranolol, diazepam, norethindrone, or ethinyl estradiol was given with 18 grams of olestra or triglyceride. There were no differences in the absorption of any of the drugs in the two vehicles, as determined from the area under the serum concentration versus time curve. In another study (32), the oral contraceptives norgestrel



398

and ethinyl estradiol were given daily over 28 days with a diet containing 18 g/day of olestra or triglyceride in a cross-over study involving 28 subjects. There were no significant differences between the two treatment periods in peak blood level, time to reach peak blood level, or area under the serum concentration-time curve for either drug. These drugs are among the most lipophilic oral medications, but are several orders of magnitude less fat-soluble than vitamin D , for example. Therefore, it is concluded that olestra will not affect absorption of even the most lipophilic drugs. Conclusion In conclusion, the safety evaluation of olestra is far broader than that normally conducted for a typical food additive. As in any safety evaluation, the animal and clinical studies conducted must be tailored to address the scientific questions posed by the chemical and biological properties of the material and the proposed uses. The current Food Additive approval process under the Food, Drug and Cosmetic Act allows for the flexibility needed to assure safety for the public and to allow innovations that will meet the needs of consumers. Literature Cited 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

Mattson, F.H.; Volpenhein, R.A. J. Lipid Res. 1972, 13, 325-28. Mattson, F.H.; Nolen, G.A. J. Nutr. 1972, 102, 1171-76. Mattson, F.H.; Volpenhein, R.A. J. Nutr. 1972, 102, 1177-80. Fallat, R.W.; Gluek, C.J.; Lutmer, R.; Mattson, F.H. Am. J.Clin. Nutr. 1976, 29, 120415. Mattson, F.H.; Jandacek, R.J. Lipids, In Press. Jandacek, R.J.; Holcombe, B.N. Lipids, In Press. Wood, F.E.; DeMark, B.R.; Hollenbach, E.J.; Sargent, M.C.; Triebwasser, K.C. Fd. Chem. Toxic.1991,29,231-36. Wood, F.E.; Tierney, W.J.; Knezevich, A.L.; Bolte, H.F.; Maurer, J.K.; Brace, R.D. Fd. Chem. Toxic. 1991, 29, 223-30. Miller, K.W.; Wood, F.E.; Stuard, S.B.; Alden, C.L. Fd. Chem.Toxic. In Press. Adams, M.R.; McMahan, M.R.; Mattson,F.H.; Clarkson,T.B. Proc. Soc. Exp. Biol. Med. 1981, 167, 346-53. Miller, K.W; Long, P.H. Fd. Chem. Toxic. 1990, 28, 307-15. Skare, K.L.; Skare, J.A.; Thompson, E.D. Fd. Chem. Toxic. 1990, 28, 69-73. Nolen, G.A.; Wood, F.E.; Dierckman, T.A. Fd. Chem. Toxic. 1987, 25, 1-8. Cortot, Α.; Phillips, S.F.; Malagelada, J.R. Gastroenterology1982,82,877-81. Hager, M.H.; Schneeman, B.A. J. Nutr. 1986, 116, 2372-77. Jandacek, R.J. J. Drug Metab. Rev. 1982, 13, 695-714. Glueck,C.J.;Jandacek, R.J.; Subiah, M.T.R.;Gallon, L.; Yunker, R.; Allen, C.; Hogg, E.; Laskarzewski, P.M. Am. J. Clin. Nutr. 1980, 33, 2177-80. Mellies,M.J.; Vitale, C.; Jandacek, R.J.; Lamkin, G.E.; Glueck, C.J. Am. Clin. Nutr. 1985, 41, 1-12.


32. BERGHOLZ


19.

399

Mellies, M.J.; Jandacek, R.J.; Taulbee, J.D.; Tewskbury, M.B.; et al. Am. J. Clin. Nutr. 1983,37,339-46. 20. Glueck, C.J.; Jandacek, R.J.; et al Am. J. Clin. Nutr. 1983,37,347-54. 21. Grundy, S.M.; Anastasia, J.V.; Kesaniemi, Y.A.; Abrams, J. Am. J. Clin. Nutr. 1986, 44, 620-29. 22. Nuck,Β.Α.;Federle, T.W. In Abstracts of the Annual Meeting of the American Society forMicrobiology;American Society for Micribiology: Washington, 1990, p. 275. 23. Mattson,F.H.;Jandacek, R.J.; Webb, M.R. J. Nutr. 1976, 106, 747-52. 24. Mattson,F.H.;Hollenbach, E.J.; Kuehlthau, C.M. J. Nutr. 1979, 109, 1688-93. 25. Jandacek, R.J.; Mattson, F.H.; McNeely, S.; Gallon, L.; Yunker, R.; Glueck, C.J. Am. J. Clin. Nutr. 1980,33,251-59. 26. Jandacek, R.J.; Ramirez, M.M.; Crouse, J.R. Metabolism 1990, 39, 848-52. 27. Koonsvitsky, B.P.; Jones, D.Y.; Berry, D.A.; Jones, M.B. Am. J. Clin. Nutr. 1991, 53, 21. 28. Suttie, J.W.; Mummah-Schendel, B.S.; Shah, D.V.; Greger, J.L. J. Am. Clin. Nutr. 1989, 47, 299-304. 29. Jones,D.Y.;Miller, K.W.; Koonsvitsky, B.P.; Ebert, M.L.;Lin,P.Y.T.;Jones, M.B.; Will,B.H.;Suttie, J.W. Am. J. Clin. Nutr. 1991, 53, 943-6. RECEIVED August 15, 1991


Safety Evaluation of Olestra - ACS Publications - American Chemical

Recommend Documents