TH ESlS I11111111111111111111111111111111111111111111111111111 3 1293 10469 This is to certify that the thesis entitled Oxidation of Cholesterol in Heated Tallow presented by Thomas Clavin Ryan has been accepted towards fulfillment of the requirements for Masters degreein Food Science Z/ éw Major pr“ essor U 2/25/91 Date 0-7639 RETURNING MATERIALS: MSU Place in 500E drop to remove this checkout from LIBRARIES your record. FINES will be charged if book is returned after the date stamped below. M ram": 25 3 . . 31‘? ‘JA§ % 35 1992 tlwivinnl‘ A,Az 517$71 . 1 ;\ <3 // ’ OXIDATION OF CHOLESTEROL IN HEATED TALLON By Thomas Clavin Ryan A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1982 ABSTRACT Oxidation of Cholesterol in Heated Tallow By Thomas Clavin Ryan The oxidative stability of tallow cholesterol and triglycerides was investigated under various processing conditions including heat cycling, frying and add-back procedures. Intermittent heating was found to be more damaging to tallow constituents than was continuous heating. French fried potatoes heated in tallow exhibited a preferential absorption for cholesterol and cholesterol oxidation products and slowed down the rate of tallow degradation. Cholesterol oxidation products isolated from heated tallow systems were identified by chromatographic means as S-cholesten-BB, 7a-diol, 5-cholesten-3B,7B-diol, 5-cholesten 3B -ol and 3,5-cholestadiene-7-one, with trace amounts of S-cholestan-BB ,Sa , 6 B-triol observed. Analysis of french fried potatoes obtained from a fast-food franchise utilizing tallow as a frying medium revealed the presence of 5-cholesten-38,7u -diol, S-cholesten-3B , 78 -diol, 5-cholesten-3B -ol, and 3,5-cholestadiene-7-one. Results of this study indicate that cholesterol readily oxidizes under frying conditions and that oxidation products are preferentially absorbed by foods fried therein. Several cholesterol oxidation products were found in commercially prepared fried foods. To my family, Mr. and Mrs. Thomas J. Ryan, Debra L. Ryan-Smith, Cynthia M. Ryan and Timothy K. Ryan ii ACKNOWLEDGMENTS The author wishes to express his sincere gratitude and appreciation to Dr. J.I. Gray for his unlimited input as research chairman and major professor during the course of this study. The patience, encouragement, and insight contributed by Dr. Gray was invaluable in the completion of this thesis. Appreciation and gratitude are also extended to the members of the guidance committee, Dr. L.R. Dugan, Dr. A.M. Pearson, Dr. M.E. Zabik and Dr. B.R. Harte for their critical review of this dissertation. The author extends his heartfelt gratitude and admiration to Dr. I.D. Morton, Chairman, Department of Food Science, Queen Elizabeth College, London, for the critical role he played in the initiation of this study and the professional influence which he left on his return to England. The author is also indebted to several individuals for their assistance in this research, including Arun Mandagere and Susan Cuppett for their help in the various areas of instrumentation employed, and Julie Luby for her aid in various procedures and the literature search. The author also wishes to acknowledge Mrs. Loraine Stephens for her assistance in typing this thesis. . Last, the author is especially appreciative to Teresa Kalil, for her support and understanding during the completion of this thesis. TABLE OF CONTENTS Page LIST OF TABLES ..... . .................... vi LIST OF FIGURES. ..... . .................... vii INTRODUCTION ........................... 1 REVIEW OF LITERATURE ....................... 3 Cholesterol Oxidation ..................... 4 Description of Cholesterol ................. 4 Autoxidation of Cholesterol ....... . . . . ...... 4 Biological Effects of Cholesterol Oxidation Products ..... ll Angiotoxic and Atherosclerotic Effects ........... ll Carcinogenic Effects .................... 16 Cholesterol Oxidation Products in Foods. . . . ........ l8 Heated Fats and Oils ..................... 24 Chemical Reactions ..................... 24 A. Oxidative Reactions .................. 24 8. Thermal Polymerization ............... . 29 C. Hydrolytic Reactions ................. 3l Effects of Processing Variables on Frying Media Stability. . . 32 A. Fat Turnover ..................... 32 8. Temperature Cycling. ................. 34 C. Heating Systems .................... 34 D. Substrate Effects ................... 35 Biological Aspects of Used Frying Oils ............ 35 MATERIALS AND METHODS ...................... 39 Beef Tallow .......................... 39 Standard Cholesterol Oxidation Products ............ 39 Chromatography Materials and Chemicals ............ 39 French Fry Potatoes ...................... 40 Commercial French Fried Potatoes ............... 40 Analysis of Continuously and Intermittently Heated Tallow. . . 41 Viscosity. ......................... 41 Color. . . ......................... 41 Iodine Value ........................ 42 Peroxide Value ....................... 42 Free Fatty Acid Value .................... 42 Fatty Acid Analysis. . ................... 42 Extraction of Cholesterol and Cholesterol Oxidation Products from Tallow .................... 42 iv Page Thin Layer Chromatography of Cholesterol Extracts ...... _43 Gas Chromatography of Cholesterol Oxidative Products ..... 43 Analysis of Tallow and French Fried Potatoes Used in Frying . . 44 Tallow Analysis . . .-. .......... . . . . . . . . 45 Analysis of French Fried Potatoes .............. 45 Fat Analysis ....................... 45 Fatty Acid Analysis .................... 45 Extraction of Cholesterol and Cholesterol Oxidation Products .................... 46 TLC of Cholesterol Extracts ................ 45 GLC of Cholesterol Extracts from Tallow Fried French Fried Potatoes. . . . ............ 45 Analysis of Commercial French Fried Potatoes .......... 45 RESULTS AND DISCUSSION . ..................... 47 The Effects of Continuous and Intermittent Heating on the Oxidative Stability of Tallow ............. 47 Effects on Tallow Triglycerides ............. 47 Effects of Heating on the Stability of Cholesterol . . . . 59 The Effects of Intermittent Frying on the Oxidation Stability of Tallow ...................... 72 Effects of Frying in Tallow ................ 72 Analysis of French Fried Potatoes ............. 31 The Occurrence of Cholesterol Oxidation Products in Commercially Produced French Fried Potatoes .......... 91 SUMMARY AND CONCLUSIONS ...................... 103 PROPOSALS FOR FURTHER RESEARCH. .................. 105 BIBLIOGRAPHY ............................ 103 LIST OF TABLES Page Oxidized sterols in foodstuffs. . . . . . ....... 21 Cholesterol content of various fats and oils. . . . . . 23 . Chemical and physical characteristics of continuously ,and intermittently heated tallow. . . . . . . . . . . . 50 Hunter calorimeter data of continuously and intermittently heated ta110w O O O O O O O O O O O O O O O O O O O O O O 58 Fatty acid analysis of intermittently and continuously heated tallow . . . . . . . . . . . . . . . . . . . . . . 6O Oxidative products of cholesterol extracted from heated tallow fractions - TLC analysis ..... . . . . . 61 Two dimensional TLC analysis of oxidation products of cholesterol in tallow heated intermittently for 75 hours and standard compounds. . . . . . . . . . . . . . 63 Physical and chemical characteristics of tallow used in frying study .......... . ........ 74 Hunter colorimeter data for tallow utilized for frying. ....................... 80 TLC analysis of the non-saponifiable fractions of tallow used in frying ................. 80 Absorption of fat by french fried potatoes fried in tallow . . . . . . . . . . . ........ . 84 TLC analysis of the non-saponifiable fraction of fat extracted from tallow-fried french fried potatoes ..... . . . . . . . . . . . . . . ..... 84 Fatty acid analysis of tallow and lipid extract from raw french fry potatoes . . . . . . . . . . . . . . 90 Fat absorption data of commercially fried french fry potatoes . . . . . . . . . . . . . . . . . . . . . . 96 TLC analysis of the non-saponifiable fraction of fat extracted from commercial french fry potatoes. . . . 99 vi Figure 10. ll. LIST OF FIGURES Page Oxidative mechanisms of cholesterol . . ......... 8 Effect of heating time on the iodine value of tallow heated intermittently (B) and continuously (C) . . 52 Effect of heating time on the viscosity of tallow heated intermittently (B) and continuously (C). . . . . . 54 Gas chromatograms of tallow non-saponifiable fractions heated intermittently for (A) 0 hours, (B) 50 hours (C) lOO hours and (D) l50 hours . . . . . . . . . . . . . 65 Gas chromatograms of tallow non-saponifiable fractions heated continuously for (A) 50 hours, (B) 100 hours and (C) l50 hours . . .................. 67 Effect of heating time on the iodinevalue of tallow heated intermittently (B), continuously (C) and intermittently with frying (D). . . ......... 75 Effect of heating time on the viscosity of tallow heated intermittently (B), continuously (C) and intermittently with frying (D) ........... 77 Gas chromatograms of tallow non-saponifiable fractions from tallow samples intermittently heated for frying. (A) 50 hours, (B) lOO hours and (C) 150 hours ..... . ............. 82 Gas chromatograms of lipid non-saponifiables extracted from french fried potatoes fried in tallow heated intermittently for (B) 50 hours, (C) lOO hours, (D) lSO hours and (A) unfried control ........ . . . . . . . ......... 85 Gas chromatograms of fatty acid methylesters from a lipid extract of unfried french fry - potatoes. . . . . . . . . . . . . . . . . ........ 88 Gas chromatogram of the non-saponifiable fraction from unused commercial frying medium. . . . . . . . . . . 92 vii Figures 12. 13. l4. Page Gas chromatogram of the non-saponifiable fraction from discarded commercial frying medium . . . . . . . . . . 94 Lipid extracts from french fried potatoes obtained from (A) an East Lansing area fast-food restaurant and (B) an Okemos area fast-food restaurant . . . . . Gas chromatograms of lipid non-saponifiables extracted from commercial french fried potatoes, (A) East Lansing area (B) Okemos area.. . . . viii INTRODUCTION Recent studies have indicated that when cholesterol is heated in air it readily undergoes oxidation to form products which may be angiotoxic (Imai et al. 1976; Peng et al., 1978) and/or carcinogenic (Smith and Kulig, T975; Bischoff, T969). Cholesterol is present in food systems of animal origin, many of which are subjected to conditions which may be conducive to the oxidation of cholesterol. Beef tallow, containing 0.15 to 0.20% cholesterol by weight (Punwar and Derse, 1978) is commonly used as a frying medium in commercial frying operations. The elevated temperature over the extended periods of time to which tallow is subjected may be compatible with the formation of oxidation products from cholesterol. In addition to the oxidation of cholesterol under frying conditions, tallow triglycerides undergo autoxidation, thermal oxidative, therma1 polymerization, and hydrolytic changes (Porkorny, l980; Perkins, 1967). The products of these reactions have been suspected of introducing potentially hazardous compounds into the diet via fried foods, as evidenced by an overwhelming number of studies reviewed by Artman (1069). The major objective of this study was to evaluate the oxidative changes in tallow triglycerides and cholesterol during exposure to frying conditions. Processing parameters such as temperature cycling, frying, and add-back procedures were utilized to monitor their effects on the oxidation of the tallow system. Various physical and chemical analysis and chromatographic techniques were employed to assess the oxidative changes in tallow constituents exposed to the various processing parameters. A second objective of this study was to determine the presence of cholesterol oxidation products in commercially processed foods, such as french fried potatoes. REVIEW OF LITERATURE Heated fats and oils, especially those used in deep-fat frying processes, have generally been regarded as having the poorest nutritional value of any edible lipids (Melnick, l957). Frying processes subject these fats and oils to high temperatures and adequate oxygen concentrations thus promoting physical and chemical changes at sites of unsaturation in the lipid moiety. The compounds formed under frying conditions not only decrease the effectiveness of the fat or oil as a frying medium, but may also pose as hazards to human health. Animal fats used as frying media may present an additional hazard due to the inherent presence of cholesterol. Recent studies indicate that cholesterol in the presence of oxygen, spontaneously forms products, some of which have been implicated as angiotoxins, carcinogens, and mutagens (Smith, 1980). Thus the review of pertinent literature will approach both the chemistry and toxicology of cholesterol and its oxidation products as well as the chemical and physical changes in frying fats and oils and the toxicological aspects of products formed during these changes. Cholesterol Oxidation Description of Cholesterol Cholesterol (S-cholesten-EB -ol) is a C27 simple neutral lipid composed of a cyclopentanophenanthrene ring system, an eight carbon side chain, attached to the O ring at theC17 position, having two tertiary carbons, and methylated at the C10 and C13 positions. The alcoholic function is at the C3 position and an unsaturated bond exists at the C5-C6 position. Cholesterol is distributed in all cells of animal tissues, with higher concentrations found in nervous tissue. Cholesterol is an integral component in cell membranes and may function as a precursor to bile acids and various steroid hormones. Autoxidation of Cholesterol Cholesterol is an unsaturated lipid and is therefore susceptible to oxidation by free radical processes via reaction mechanisms analagous to those involved in the autoxidation of unsaturated fatty acids (Smith, T980). The sensitivity of cholesterol to oxygen has been noted in various reviews. Smith (1980) reported that Schulze and Winterstein investigated the sensitivity of crystalline cholesterol to air as early as 1902. Bladon (1958) mentioned early studies by Lifschultz in 1°07 describing the isolation of ”oxycholesterol," believed to be a pure chemical species formed in cholesterol which had been exposed to air. Bergstrom and Samuelsson (1961) cited the work of Lamb (1914) implicating oxygen as a necessary reactant for oxidative changes in cholesterol colloidally dispersed in an aqueous system. Bischoff (1930) confirmed these findings by showing that no "oxycholesterol" formed from cholesterol in the absence of air. A series of experiments were conducted by Bergstrom and Hinterstein from 1941-45 which were later reviewed by Bergstrom and Samuelsson (1961). Molecular oxygen was utilized as the oxidizing species in a number of experiments on aqueous colloidal dispersions of cholesterol. "Oxycholesterol" was found in this oxidizing system. Crystallization and chromatographic techniques revealed that aeration of cholesterol for five hours at 85°C caused an eighty percent conversion of cholesterol to "oxycholesterol." Furthermore, "oxycholesterol" was found to contain S-cholesten-38, 7a-diol, S-cholesten-BB, 7B -diol, 5- cholesten-3B -ol-7-one, and 3,5-cholestadiene-7-one. These studies also showed that prooxidant ions such as cupric and ferric species catalyzed the oxidative reaction. Bergstrom and Hinterstein postulated an oxidative mechanism based on free radical processes whereby molecular oxygen attack at the C7 position yields an unstable hydroperoxide. They attributed the presence of 5-cholesten-3B-ol-7-one to the dehydration of the 7-hydr0peroxide, and the epimeric 3,7-diols to form directly from the simple decomposition of the 7-hydroperoxide. The epimeric nature of these compounds was believed to be from either two stereoisomeric hydroperoxides, or due to rearrangement from a common precursor during decomposition. Trace amounts of cholestan-38, 5a,68-triol and 3,5-cholestadiene-7-one reinforced the validity of this proposed scheme . Bladon (1958) also recognized the free radical induced nature of the commonly formed oxidation products. A similar mechanism was thus proposed with a 7-hydr0peroxide as the primary oxidation intermediate. However, Bladon (1958) theorized that the epimeric 3,7-diols commonly formed were the product of the reaction of a 7-hydroperoxide with a propagating species such as an unoxidized cholesterol molecule. Dehydration and oxidation of these compounds accounted for the various secondary oxidation products found in autoxidized cholesterol. The elucidation of the reaction mechanism involved in sterol autoxidation was greatly assisted by application of thin-layer chromatographic techniques (Smith, 1980). Analysis of a 12-year old sample of originally pure cholesterol by two-dimensional thin-layer chromatography revealed the presence of more than 30 distinct compounds including 3,5-cholestadiene-7-one, 5-cholesten~38 -ol-7-one, cholestan 38 ,Su ,6B-triol, and the epimeric 5 cholestan-38,7a-diol and 5- cholesten-3B,7B-diol (Smith et al., 1967). Confirmation of the reaction mechanism was hindered due to the inability to isolate 7-hydroperoxycholesterol, which was the proposed primary oxidation product. Van Lier and Smith (19706) induced hydroperoxide formation by heating thin layers of cholesterol in air for 7 days at 100°C. A series of fractionations and recrystallizations produced a fraction containing a S-hydroperoxide which underwent complete isomerization to 7-hydroperoxycholesterol during Sephadex LH-20 separation. The identity of this compound was confirmed by gas chromatography. A specific and sensitive thin-layer chromatographic technique was developed by Smith and Hill (1972) to identify sterol hydroperoxides. N,N-Dimethyl-p-phenylenediamine and N,N,N',N'-tetramethyl-p-phenyl- enediamine dihydrochlorides were utilized as visualizing agents, yielding characteristic colors upon application. Autoxidized cholesterol samples were found to contain 7 -hydroperoxides as the primary autoxidation products. The overall autoxidation scheme for cholesterol was reviewed by Smith (1980). It was generally recognized that two distinct oxygen dependent processes occur in autoxidation, with hydroperoxides representing the primary products. 02 + RH—DROOH (1) RCH(OH)+O2 —b RC=O + H202 (2) The major reactions and products formed are shown in Figure l. The initial reaction involves the abstraction of the allylic C7 hydrogen and reaction with ground state dioxygen to form a C7 peroxy radical. This peroxy radical is stabilized by hydrogen abstraction to form the epimeric 7wumuwxo .P «Lama; 335335-113-3 4.33.5.8 L... . a uzo-s-uzuafimu-azu -9 n uoncumoco>z unh-JO-nn uzuhau-uozo :30 u 4 Y? 403.... .- Q6 .10 in 2559.010 .0 402.710 .Qodn 2 ROOH + R° ROOH _. R0- +-0H Termination R0 + R° ---DIR R R- + R000 —> -ROOR R000 + R000 —-> ROOR + 02 RH refers to any unsaturated fatty acid in which the H is labile by reason of being on a carbon atom adjacent to a double bond. R- refers to a free radical formed by removal of a labile hydrogen. The onset of this reaction mechanism is characterized by absorption of oxygen by the fat. During this induction period, antioxidant species in the fat are consumed while free radicals or their precursors accumulate. Rapid oxygen uptake proceeds at elevated radical concentrations, evidenced by increased peroxide values. Various researchers have noted decreased induction periods and increased rates of oxidation under certain conditions including increased temperature (Gunstone and Hilditch, 1945), increased surface-volume ratios (Pokorny, 1966), increased unsaturation and addition of preformed hydroperoxides to unoxidized oils (Gunstone and Hilditch, 1945). Artman (1969) has comprehensively reviewed the autoxidation process. Radical abstraction of an allylic hydrogen from an unsaturated fatty acid forms a fatty acyl radical. Direct addition of 26 atmospheric oxygen, or addition after double bond rearrangement, yields a peroxy radical which abstracts a new allylic hydrogen from an unoxidized unsaturated fatty acid. The resulting hydroperoxide may then undergo homolytic cleavage, forming two new radicals which induce an autocatalytic propagation of the autoxidation mechanism. Termination reactions occur when the concentration of radicals are sufficiently great to allow kinetically preferential radical-radical interaction. Products of autoxidation, although dependent on the type of oxidized fat or oil, generally include short chain aldehydes, ketones, alcohols and acids which render the oil aesthetically unacceptable. Trace amounts of esters, hydrocarbons, aromatics, cyclohexanes and lactones have also been isolated from autoxidized samples (Artman, 1969). Dimeric and polymeric compounds also form in autoxidized lipids at the expense of unsaturated bonds. The formation of these compounds is concommitant with decreases in iodine value and increases in viscosity of the oil. Conflicting reports of the chemical nature of these polymers have been cited in the literature. Artman (1969) stated that both carbon-carbon linked polymers and carbon-oxygen linked polymers have been isolated from autoxidized lipid systems. O'Neill (1954) isolated ether linked dimers from oleate, but found only carbon linked species in linoleate. Chang and Kummerow (1953) cleaved methyl oleate polymers with strong acids, thus indicating ether or peroxy linkages. Kartha (1960) observed oxygen linkages in polymerized methyl oleate held at 25°C while carbon linkages predominated under similar conditions at 60°C. Triolein and methyl oleate exhibited completely different polymeric products when oxidized under similar conditions. 27 Autoxidation reactions are involved in the degradation of frying media held at ambient temperatures. Artman (1969) described the susceptibility of intermittently heated fats and oils to autoxidative reactions. Formation of hydroperoxides occurs more readily in thermally damaged oil. Buildup of hydroperoxides during cooling cycles of frying fats, followed by reheating, accelerates the liberation of radicals and subsequent degradation of the fat at higher temperatures. 2. Thermal oxidation Oxidation of a fat or oil in the presence of oxygen at temperatures in excess of 100°C is referred to as thermal oxidation (Perkins, 1967). Temperature, time, rate of aeration, and time-temperature cycling influence the rate of thermal oxidative reactions. It is generally accepted that oxidative reactions at elevated temperatures proceed via reactions similar to autoxidation; however, the transient nature of hydroperoxides at frying temperatures suggests the development of other primary oxidation products. Artman (1969) noted differences in both reaction rate and type of products between thermal oxidation and autoxidation. Furthermore, Artman postulated direct radical formation at elevated temperatures without the intermediacy of hydrOperoxides. Fats and oils, when heated to high temperatures, absorb oxygen thus increasing the peroxide value of fat or oil systems. Further beating causes the decomposition of peroxides, accompanied by loss of weight due to volatilization of short chain scission products (Artman, 1969). Double bonds conjugate and oxygenated products accumulate as evidenced by increased ultraviolet absorption. The oxygenated products include carbonyls, hydroxy, and epoxy derivatives (Perkins, 1967). Hith continued heating, conjugation of unsaturated fatty acids such as 28 linoleic acid occurs. The levels of conjugated acids increase, reach a maximum, then decrease as various polymerization and decomposition reactions take place. This phenomenon is represented by decreasing iodine values with increased time of heating. Browning occurs in the fat and can be attributed to a,8-and o, of-unsaturated carbonyl compounds as well as highly polar dimeric compounds (Perkins and Kummerow, 1959). Oxidative polymerization reactions increase the viscosity of the fat or oil, seriously deteriorating the foaming stability. Investigations of the higher molecular weight products formed in heated fats indicate a variety of compounds. Frankel et al. (1960) isolated carbon-linked dimers which were acyclic in nature, some of which contained polar functional groups. Micheal, (1966a and b) reported both polar and non-polar dimers. Non-polar dimers were characterized as Diels-Alder adducts with six-membered unsaturated rings, much like those formed in anaerobically heated fats. Polar dimers included both carbon and oxygen linked dimers with polar functional groups; These were found to be responsible for the increased foaming tendencies of thermally oxidized oils. Melnick (1957a) stated that oxidative polymers cause serious flavor problems in frying fats and oils. Many products formed in thermal oxidative reactions have also been identified in lipid systems undergoing autoxidation (Perkins, 1967). Low molecular weight alkanes, saturated and unsaturated aldehydes, ketones and alcohols have been isolated from the volatile fractions of various thermally oxidized fats and oils. In addition, pyrolysis products such as methylcyclopentanes, cyclohexanes, benzene derivatives, acrolein, water, carbon monoxide and carbon dioxide have also been reported (Artman, 1969). A wide variety of compounds formed 29 via conjugation and trans isomerization have also been isolated and identified, indicating the complex reaction processes occurring in heated fats. Several researchers has suspected the presence of cyclic monomers in thermally oxidized fats and oils. Firestone et al. (1961) heated cottonseed oil and found a fraction which did not adduct with urea. Micheal et al. (1966) oxidized methyl linoleate at 200°C hours and identified aromatic as well as alicyclic monomers in the system. Artman and Alexander (1968) found alicyclic, cyclic, and aromatic compounds in thermally oxidized partially hydrogenated soybean oil. Micro-constituents of fats and oils are also susceptible to oxidative changes at elevated temperatures. Non-saponifiable compounds such as sterols and carotenoids show marked changes as a consequence of heating (Artman, 1969). It has been suggested that sterols may oxidize to volatile or polymeric species when thermally abused. Fioriti and Sims (1967) oxidized cholesterol for nine weeks at 82°C and documented an 80% conversion to oxidation products including epoxy, ketone, and polyhydroxy derivatives. B. Thermal Polymerization Anaerobic thermal polymerization reactions are those occurring in fats and oils heated in excess of 100°C in the absence of air (Perkins, 1967). The prevalence of purely thermal reactions in frying fats is unknown, yet conditions such as low oxygen concentrations in frying fats and occulsion of air by steam generated from the fried product may facilitate these reactions. Consequently, various heating studies on fats and oils have been conducted in vacuo or under inert atmospheres. Roth and Rock (1964a) studied anaerobically heated fats, finding little change after prolonged exposure to frying temperatures. Results of various studies indicate that the development of toxic 30 compounds in thermally abused fats and oils require much higher temperatures than those used in frying (Artman, 1969). Thermal polymers form via Diels-Alder reactions between unsaturated. fatty acyl groups after conjugation has taken place (Perkins, 1967). These reactions proceed at the expense of non-conjugated double bonds in the triglyceride moiety, as indicated by decreased iodine values and increased viscosity levels as heating time progresses (Sims, 1957; Smouse, 1975). These polymerization reactions may occur between two triglycerides by intermolecular dimerization or by reaction between unsaturated fatty acyl groups on the same triglyceride molecule, referred to as intramolecular dimerization. Transesterification of the latter yields higher molecular weight polymers similar to those formed via the intermolecular process (Artman, 1969). Many studies have been carried out on the thermal polymerization of fats. Perkins (1967) described thermally induced polymerization as it involves linoleic acid content in the triglycerides. Conjugation of the double bonds occurs in the presence of heat. The conjugated species then may react with another unsaturated fatty acyl group or undergo an intramolecular cyclization. This results in a six-membered cyclic compound of either monomeric or dimeric nature, respectively. These polymeric intermediates contain one endocyclic double bond and one exocyclic double bond, the latter possessing the ability to act as a dienophile. Trimerization reactions may then occur, leading to high molecular weight polymeric compounds. Other researchers have isolated similar compounds in thermally oxidized fats. Evans et al. (1965) isolated Diels-Alder products from thermally abused fats, as well as non-cyclic dehydropolymers and hybrid dimers in which one monomeric unit had cyclized prior to dimerization. Sahasrabudhe (1965) found 31 1,2-disubstituted cyclohexanes in anaerobically heated polyunsaturated fatty acids, indicating Dials-Alder mechanisms at work. In addition to polymeric substances, volatile and monomeric species have also been isolated and characterized from thermally abused fats and oils. Volatile constituents include typical pyrolysis products of lipids such as short chain aldehydes, ketones, olefins, carbon monoxide, carbon dioxide, and water (Artman, 1969). Many of these compounds are highly reactive and may become incorporated in concommitantly formed polymers. Non-volatile monomers isolated from thermally abused fats include free fatty acids liberated in transesterification reactions. Long chain ketones may result via decarboxylative dimerization of these fatty acids. Acrolein is commonly identified as a pyrolysis product of heated fats, formed from the glycerol moiety (Artman, 1969). Concern over the presence of thermal polymers in frying fats began when Crampton et al. (1953) reported that they suppress growth and exhibit toxic effects in rats. Melnick (1957a) regarded thermally induced polymers as the inherent danger in fried foods. This was due to the fact that thermal polymers do not exhibit the strong objectionable odors and flavors associated with oxidative polymers. To the contrary, Melnick (1957a) stated that thermal polymers enhance the flavor stability of frying oils. Thus the safeguard of unpalatabilitv is negated when thermal polymerization reactions predominate in heated fats and oils. C. Hydrolytic Reactions Hydrolysis of heated fats liberate free fatty acids via cleavage of esterified fatty acyl groups from triglyceride molecules in the presence of water. These long chain fatty acids as well as the 32 resultant mono-and diglycerides limit the functional life of a frying fat by reducing the smoke point (Artman, 1969). Hydrolysis of triglycerides has also been associated with increased foaming in fats used in frying operations (Smouse, 1975). The presence of water in fats, whether introduced through the frying of foods or as a product of oxidative reactions, facilitates hydrolytic reactions in heated fats (Artman, 1969; Pokorny, 1980). The accelerated deterioration of fats by the addition of water is well documented in the literature. Perkins and Van Akkeren (1965) injected water into heated fats and found more deterioration than encountered in similar fats which were utilized in frying operations. Yukui (1965) whose research was reviewed by Artman (1969), noted the strong hydrolytic action of water in heated fats. It was found that water caused both hydrolytic and oxidative deteriorations, the extent depending on temperature and degree of exposure to air. Hydrolytic reactions predominate when heated fats are exposed to water under low oxygen atmosphere. At elevated temperatures the effect of water is minimized due to its rapid volatilization. Used frying fats exhibit elevated free fatty acid levels with average values being around 1% (Bates, 1952). It was important to note that acid values not only reflect the hydrolytic activity in an oil, but also oxidative changes since various scissionings of dismutated hydroperoxides liberate fatty acyl compounds having carboxyl functions (Smouse, 1975). Effects of Processing Variables on Frying_Media Stability A. Fat Turnover The loss of frying medium via absorption and the resultant replenishment with fresh oil is referred to as fat turnover. This 33 process maintains the quality of fats utilized in large volume continuous frying operations by both increasing the pr0portion of fresh fats in the frying medium and by the absorption of deleterious compounds into the fried product. Constant rates of fat removal and replenishment result in a steady-state condition of the fat, which does not deteriorate extensively with continued use. Factors such as size of frying loads, frequency of frying, product, and product dimension influence the fat turnover rate, and consequently the life of the frying fat. High frying loads increase the frequency of add-back to the fryer. Use of high frying loads in a continuous process allows steady-state conditions to develop rapidly (Melnick, 1957b; Perkins and Van Akkeren, 1965). Prolonged use of commercial frying media have been reported under these conditions (Artman, 1969). Product constitutent differences exhibit a profound effect on turnover rate of a frying fat (Pokorny, 1980). Foods which absorb high levels of fat during frying facilitate more frequent or substantial levels of oil add-back. Similar products having different surface areas demonstrate differences in absorption rates. In potato frying operations, french fried potatoes contain approximately 13-14% fat after frying (Watt and Merrill, 1963). Fried sliced potatoes of 12 mm thickness contain 7% fat (Strock et al., 1966) while 1 mm slice contain 30-40% fat as cited by Artman, 1969. Consequently, oils used in potato chip frying operations exhibit extended frying life. Melnick et al. (1958) surveyed oil samples from various chippers and found only slight changes between used oils and the corresponding fresh oils. This stability of used oils was attributed to the high rate of fat turnover. Non-continuous frying operations, mainly encountered in restaurants or fast food establishments, expose fats to extended periods of time at 34 high temperature, with small and infrequent frying loads. The decreased stability and functional life of such oils clearly demonstrates the protective effect of high turnover rates (Artman, 1969). B. Temperature Cycling Melnick (1957b) noted the greater stability of frying oils used in continuous operations, compared to those used intermittently. This effect was studied by Perkins and Van Akkeren (1965) by heating cottonseed oil both continuously and intermittently. The intermittently heated samples were found to decompose more rapidly than the continuously heated controls. The decreased stability of intermittently heated samples was attributed to buildup of fatty acyl peroxides on the surface of the oil at ambient temperatures, followed by reheating. At elevated temperatures, the peroxides decompose, damaging the oil via liberation of hydroxyl and carbonyl compounds. This damage, estimated by hexane insolubles concentration or amount of polar material present, was equivalent after 62 hours of intermittent heating and 166 hours of continuous heating. C. Heating Systems Intense localized heating systems, such as high temperature heating elements, may influence the deterioration of a fat or oil used in frying. Such elements not only expose part of the fat to extremely high temperatures, but also increase agitation of the fat via convection currents (Artman, 1969). Circulating systems utilizing heat exchangers eliminate localized heating, but generally increase - agitation and thus the aeration of a fat (Roth and Rock, 1964b). 35 D. Substrate Effects Compositions of fried substrates may influence the stability of frying oils in various ways (Pokorny, 1980) 1. Via release of antioxidants or prooxidants from the substrate into the oil or by absorption of such compounds in oils by the fried substrate. 2. Via catalytic effects of various functional groups of the substrate on free radicals and secondary reactions in the frying oil. 3. Via adsorption or chemisorption of oxidation products on the fried substrate. Constituents in fried substrates which enhance oxidation in frying oils include polyunsaturated lipids and phospholipids introduced through migration and water which promotes hydrolysis reactions. Constituents possessing antioxidant activities include essential oils from herbs and spices, sterols, carotenoid pigments, and tocopherols. Biological Aspects of Used Frying Oils Many studies conducted to assess the biological hazards of heated fats and oils generally employed samples which were intentionally abused by extremes of temperature and aeration. These studies indicated that toxic compounds exhibiting adverse physiological effects on test animals were formed. However, a number of these researchers expressed the irrelevence of their results in the application to fats and oils used under normal frying conditions. Therefore, this review will address only those studies conducted under actual or simulated frying conditions. The accumulated results of studies over the last three decades indicate that fats and oils used under practical frying conditions are not harmful to test animals. Deuel et a1. (1951) fed to male and 36 female rats margarine which had been used for 10 weeks for deep fat frying, finding no adverse effects. Kaunitz et al. (1956) isolated the distillate of fat used in frying for 80 hours at 190°C. Administration of the distillate via dietary supplementation resulted in only slight decreases in both net dietary energy and growth rate Keane et al. (1959) collected fat samples used up to 24 days in commercial frying operations. The response of rats fed fat for 7 weeks at 18 percent of the diet indicated that the used fats had higher caloric efficiencies, yielding better growth rates with no evidence of toxicity. Poling et al. (1960,1962), obtained thirty-four fat samples from various frying operations. These samples caused no liver enlargement or loss of available biological value in rats on supplemented feed for 7 days. This was in great contrast to results obtained with laboratory heated fats without frying. In cottonseed oil used to fry potato chips, Rice et al. (1960) noted decreased feed intake, increased liver size, and decreased growth in rats fed a 20% fat diet, but only after long-term heating in which the fat foamed violently during frying. Artman (1969) cited the work of Kajimoto and Mukai (1964) who reported a correlation between foaming levels during frying and fat degradation. Fats exhibiting severe foaming substantially decreased growth rats of weanling rats when fed at 10 percent of their diets. Also cited by Artman (1969) was the extensive work of Mameesh et al. (1965, 1967) in which cottonseed oil, heated 8 hours daily at 195°C for 25 days, was used to fry broad bean cakes. The used oil was fed to rats at 10% supplementation in a diet of bread, milk, and salt. Unfried bean cake was fed to the control group. Lower feed intake and lower fat absorbability caused a lower growth rate in rats fed the heated fats. Also noted was lower levels of fat in the 37 livers, but liver size did not change. Autopsy results of the rats receiving fat heated for 204 hours indicated hyperplastic lymphoid tissue on the small intestine and evident changes in liver nuclei after 8 weeks of feeding. Similar results were observed in fat heated for 68 hours, but only after 17 weeks of feeding. The vast differences in these results compared to those reported by other authors who fed abused oils was attributed to the suppressive effect of frying on the formation of toxic compounds Perhaps the most comprehensive study was conducted by Nolan et al. (1967) to assess long-term effects of feeding various frying oils. Five fats were heated intermittently at 182°C and were used to fry potatoes, onions, and fish. The frying schedule and intermittent heating corresponded to the most abusive commercial practices. Four of the fats were heated until foaming persisted, rendering the fats impractical and unsafe by commercial standards. This time ranged from 49-116 hours. The fifth fat contained methyl silicone which prevented foamdng, so frying was discontinued after 216 hours. Each fat and unheated control was administered in the diet of 100 rats at 15 percent from time of weaning to 2 years of age. Rats fed heated fats exhibited slightly poorer growth rates and feed efficiencies than controls, which was attributed to the concentration of non-absorbable polymers in the (heated fats. No increase in mortality, increase in tumor incidence, or irregularities monitored by extensive biochemical or histological examination was evident. Analysis of the used fats indicated low levels of distdllable, non-urea-adductable materials. Separation, concentration, and administration of these compounds to weanling rats at extremely high levels gave adverse toxic effects. The author 38 concluded that the absence of adverse effects in the two year study clearly indicate that such substances offer no hazard in a chronic feeding situation analogous to human consumption of used frying fats. MATERIAL AND METHODS Beef Tallow A 100 gallon drum of refined edible beef tallow was obtained from a commercial supplier and stored at 4°C. The same lot of tallow was used for the duration of the study to avoid variations among tallow samples. Standard Cholesterol Oxidation Products Standard 5-cholesten-38 -ol, 5-cholesten-38, 7o -diol, 5-cholesten-38, 78-diol, 5-cholesten-38-ol-7-one, 3,5-cholestadiene-7- one, 5-cholesten-38, 25-diol, 5-cholestan-So,6-epoxy-38-ol, cholestan-38 , 5o, 68-triol, 56- cholestan-38-ol, 5 o-cholestane-B-one, So -cholestan-3,6-dione, So-cholestan-38-ol-7-one, cholestan-38, 56-diol -6-one, 5-cholesten-35, 20 -diol, 56- cholestane,5,24-cholestadiene-38 -ol, 4-cholesten-3-one, 5-cholesten-3-one, 5-cholesten-38, 4o -diol, and 5-cholestan-58 , 6-epoxy-38-ol were purchased from Steraloids Inc., NIlton, New Hampshire, USA. Chromatography Material and Chemicals Silica gel H was purchased from Merck, Darmstadt, Germany, while Redi/Plate Precoated Silica Gel GF (20 cm x 20 cm) were purchased from Fisher Inc., Pittsburgh, Pa. All other reagents and chemicals used in 39 40 the analysis were analytical grade. French Fry Potatoes Hearty-House Grade A, 3/8" Crinkle Cut Potatoes were acquired from Vita-Bite Foods, Portland, Oregon. The french fry potatoes contained; potatoes, partially hydrogenated vegetable shortening (soybean, palm and/or cottonseed oil) and/or beef fat, disodium dihydrogen pyrophosphate (to promote color retention), and dextrose. The french fry potatoes were stored at -20°C until use. Commercial French Fried Potatoes French fried potatoes samples were acquired from two commercial fast-food franchises in the East Lansing-Okemos area. These samples were stored in polyethylene bags at -20°C until required for analysis. A three phase approach was utilized to investigate the effect of frying temperatures on the oxidative stability of cholesterol in tallow. In Phase I of the study, the effect of continuous versus intermittent heating on the cholesterol stability in tallow was investigated. Phase 11 involved the analysis of both tallow and french fries produced in the laboratory under practical operational conditions, in regard to the oxidative state of cholesterol. Phase III of the study dealt with investigations of commercially produced french fried potatoes by application of techniques and methodology attained in the two previous phases, to determine the prevalence of cholesterol oxidation products in commercially prepared foods. 41 Phase I Analysis of Continuously and Intermittently-heated Tallow Two heating trials were carried out in this phase of the study. In Trial 1, tallow was heated continuously for 300 hours at 180:2°C, samples being removed for analysis every 50 hours. In Trial 11, tallow was heated intermittently for 8 hours per day, for a total heating time of 275 hours. Samples were removed for analysis every 25 hours. No tallow was added back to the fryer in this study. In both trials, 7 kg of tallow were heated in a General Electric Hotpoint Deep-Fat Fryer (Model No. HK3). Collected samples were stored at -20°C until analysis. Viscosity Viscosity of all samples was determined using a Nametre 7006 Direct Readout Viscometer in conjunction with an Exacal 100 controlled temperature circulating water bath at 501,010C. Approximately 409 of heated tallow samples were placed in a beaker and allowed to equilibrate to 50°C for 15 minutes prior to immersion of the viscometer head. Once immersed, the sample was allowed to equilibrate for 15 minutes prior to recording the viscosity. Color A model 0-25 Hunter Color Difference Meter with an inverted head was used to measure color differences in the heated tallow samples. A standardized yellow tile (L=78.4, aL=-l.9, bL=25.0) was utilized as a color reference. Approximately 709 of tallow samples were placed in an optical glass cylinder cup (7.4 cm x 1.9 cm). Color measurements were taken, aided by covering the samples with an inverted white-lined can for standard optical background. 42 Iodine Value Iodine values for all tallow samples were determined using the official AOAC Method (#28.021). Peroxide Value Peroxide values for all tallow samples were determined using the official AOAC Method (#28.023) Free Fatty Acid Value Free fatty acid values for all tallow samples were determined using the official AOAC Method (#28.030) Fatty Acid Analysis All tallow samples were methylated by the boron trifluoride- methanol method of Morrison and Smith (1964) utilizing the preparative conditions for triglycerides. Analysis of fatty acid methyl esters was carried out with a 5830A Hewlett Packard gas chromatograph using a glass column (2m x 4mm i.d.) packed with 15% diethyleneglycol succinate on Chromosorb N 80/100 mesh (Supelco Inc., Bellefonte, Pa.) The analysis was carried out isothermally at 190°C, with an injection port temperature of 210°C, and flame ionization detector temperature of 300°C. Nitrogen carrier gas flow rate at the detector was 40 m1/minute. Extraction of Cholesterol and Cholesterol Oxidation Products from Tallow The non-saponifiable fractions which contained cholesterol and any oxidative derivatives were obtained from 1009 of each tallow sample by the saponification and extraction procedure described by Itoh et al. (1973). One thousand ml of 1.0M alcoholic KOH were refluxed with the tallow sample for 1 hour followed by multiple extractions with isopropyl ether. After extraction, the non-saponifiables in solution were evaporated to dryness in a Buchi Rotovapor R rotary evaporator 43 (Buchi Inc. Switzerland), redissolved in a known volume of ethyl acetate, flushed with nitrogen and stored at -20°C until analysis. Thin-Layer Chromatoggaphy of Cholesterol Extracts Analytical thin-layer chromatography plates (20cm x 20cm) were prepared using silica gel H spread at a thickness of 0.4mm. The plates were allowed to dry overnight, followed by activation at 120°C for 2 hours. One-dimensional thin-layer chromatograms were obtained for all heated tallow non-saponifiable fractions. Similarly, selected standard cholesterol oxidation products were also run on thin-layer plates. Twenty pl aliquots of samples and standards were spotted onto the plates which were developed three times in an ethyl acetate-heptane (1:1, v/v) solvent system, with air drying between irrigations. The thin-layer plates were then Sprayed with 50% H2504 and heated for 15 minutes at 120°C for color development. Rf values and spot colors were recorded. Two dimensional thin-layer chromatography was performed on the standard cholesterol oxidation products and a representative non-saponifiable fraction from a tallow sample which had been intermittently heated for 75 hours. Twenty ul aliquots were spotted on previously described thin-layer plates. Three developments in the first dimension using ethyl acetate-heptane (1:1, v/v) was followed by two developments in the second dimension using ethyl acetate-benzene (2:1, v/v). Visualization by 50% H2504 was performed as previously described. Gas Chromatography of Cholesterol Oxidative Products Packed and capillary columns were utilized in the gas-liquid chromatographic analysis of standard cholesterol oxidation products and the non-saponifiable fractions of all heated tallow samples. Packed - 44 column gas chromatography analyses were performed using a Hewlett Packard 5830A gas chromatograph equipped with a flame ionization detector and a 2m x 4mm i.d. glass column packed with 3% SP-2100 on 100-120 mesh Supelcoport (Supelco, Inc., Bellefonte, Pa.). The chromatograph was operated isothermally at 260°C with a nitrogen flow rate of 40 ml/minutes at the detector. Temperatures of the detector and injection port were 350°C and 280°C, respectively. Capillary gas chromatographic analysis was performed using a Hewlett Packard 5840A gas chromatograph equipped with a flame ionization detector. A glass capillary column (30m) coated with SP-2100 (Supelco, Inc., Bellefonte, Pa.) was used and the chromatograph was operated isothermally at 235°C with a back pressure of 0.5 kg/mZ. Temperature of the detector and injection port was 400°C and 120°C, respectively. ' Phase II-Analysis of Tallow and French Fried Potatoes Used in Frying Seven kg of tallow were placed in a deep-fat fryer and heated to 18032°C. The tallow was heated intermittently for 12.5 hours per day for a total of 200 hours. Two 7009 samples of french fried potatoes were fried for seven minutes per each 12.5 hour heating period. The fried samples were wrapped in aluminum foil and held at -20°C until analysis. Tallow samples were collected at the end of each 25 hour heating period, and stored at -20°C until analysis. Fresh tallow was added back to the original mark in the fryer when necessary, and the entire tallow volume was filtered every other day. 45 Tallow Analysis The analysis of the tallow used to french fry potatoes was carried out as previously described with the following exceptions: 1. Fatty acid analysis was omitted due to poor resolution encountered in previous studies. 2. Sample weight for cholesterol oxidation product analysis was reduced from 1009 to 159. 3. One-dimensional thin-layer chromatography was performed using precoated Silica gel G plates (0.25mm thickness). 4. Only packed column gas chromatography was performed. Ana1ysis of French Fried Potatoes Fat Analysis The fat content of the french fried potatoes was determined by solvent extraction. A 1009 aliquot of four frying samples (representing a 25 hour period of heating) were homogenized for 3 minutes in a Virtis homogenizer (Model #45) with 150 ml of hexanes- diethyl ether (9:1 v/v). The solvent extracts were filtered and combined. The homogenate was reextracted for 2 minutes with a similar volume of solvent. Combined extracts were filtered and dried for three hours over anhydrous sodium sulfate. The solvent was removed in a Buchi rotary evaporator. The fat was weighed, flushed with nitrogen, and stored at -20°C until analysis. Unfried french fry potatoes were similarly extracted. Fatty Acid Analysis Fatty acid analysis, as described previously, was performed on lipid extracted from unfried french fry potatoes. 46 Extraction of Cholesterol and Cholesterol Oxidation Products Non-saponifiable fractions of fat absorbed by french fried potatoes were obtained by methods described earlier with the exception that the sample weight was 15.09. Thin-layer Chromatography of Cholesterol Extracts Thin-layer chromatography was performed as previously described with the exception that precoated Silica gel G plates (0.25mm thickness) were used. Gas Chromatography of Cholesterol Extracts from Tallow-Fried French Fried Potatoes Packed column gas chromatography was performed on non-saponifiable fractions of french fry lipids as described previously. Dilutions (1:1) with ethyl acetate were necessary for on-scale chromatograms. Phase III-Analysis of Commercial French Fried Potatoes French fried potato samples were obtained from two local fast-food franchises, collected every two days and stored at -20°C until analysis. Six samples covering a 12 working day period were obtained from each franchise. Fat contents, thin-layer and gas-liquid chromatographic analyses were performed as described previously in Phase II. RESULTS AND DISCUSSION Phase I - The Effects of Continuous and Intermittent Heating on the Oxidative Stability of Tallow. A. Effects on Tallow TrigLycerides The analysis of intermittent and continuously heated tallow clearly demonstrates the damaging effect of temperature cycling on the oxidative stability of frying fats. In addition, intermittently heated tallow was found to deteriorate at a much faster rate than tallow heated continuously. Perkins and Van Akkeren (1965) also noted that intermittent heating caused a more rapid decrease in the quality of cottonseed oil compared to continuous heating. This disparity in rate of deterioration can be attributed in part to the type of degradation reactions which predominate at various temperatures in the heating cycle. Frying temperatures around 180°C are conducive to thermal oxidation and polymerization reactions in both intermittent and continuously heated tallow. Exposure to air at ambient temperatures facilitates autoxidation reactions in the intermittently heated tallow as well (Perkins, 1960, 1967). Tallow as a fat has a solid crystalline structure at ambient temperatures. While in this state, surface-layer tallow triglycerides are directly exposed to autoxidative attack by atmospheric oxygen. Intermittent heating subjects tallow to such conditions. In addition, Artman (1969) stated that fats and oils previously damaged by heating are more susceptible to autoxidative reactions. 47 48 At ambient temperatures, hydroperoxides accumulate on the surface of the solid tallow matrix. Nhen reheated, these hydroperoxides decompose, resulting in the liberation of free radicals and secondary, oxidation products into the melted tallow moiety. Subsequently, these products of autoxidation may become involved in thermal oxidative or polymerization reactions at elevated temperatures. In the case of prolonged intermittent heating, this process repeats itself many times over, with autoxidation reactions occurring more readily as the total heating time increases. Likewise, the products formed in these processes accumulate in the heated tallow, resulting in increased concentrations of deleterious compounds. Fryer design may also significantly influence the deterioration rate of intermittently heated tallow (Artman, 1969). Direct exposure heating elements, like those employed in this study, generate extremely high localized temperatures. The solid matrix of tallow at ambient temperatures completely surrounds these elements. Nhen reheating occurs, those tallow triglyercides in contact with or adjacent to the element are exposed to temperatures sufficiently high as to induce thermal polymerization of unsaturated fatty acyl grOUps. This condition persists until an adequate amount of the fryer contents has melted to allow suitable mobility for convection currents in the liquid tallow to be established. Convection currents also contribute to the incorporation of oxygen in the heated tallow via entrapment at the fryer surface. Continuously heated tallow deteriorates much more slowly than intermittently heated tallow, although the degradative mechanism is similar. The major deterrent to rapid deterioration is the lack of 49 autoxidative reactions in continuous heating. Propagation of the oxidative mechanism is limited due to the constant high temperature of the heated tallow. As indicated by Artman (1969), hydroperoxides which are the primary product of autoxidation reactions, are transient at best at frying temperatures. The lability of these compounds precludes their formation of secondary oxidation products. Localized heating effects may influence the stability of continuously heated tallow only at the onset of heating. Convection currents which were present for the duration of the continuous heating trial, slowly induced incorporation of oxygen into the tallow matrix, providing adequate oxygen concentrations for thermal oxidation to occur. The results obtained in the analysis of heated tallow are presented in Table 3. Peroxide values for tallow samples heated intermittently or continuously did not exceed 3.65, reflecting very low levels of peroxides. Samples obtained for analysis were removed from the fryer while the tallow was still hot, which may explain the low values observed. Perkins (1967) noted that heated fats and oils exhibit low peroxide values, which was attributed to the aforementioned lability of such compounds at elevated temperatures. The fluctuation in peroxide values in both heating trials is attributed to this lability. A slight decreasing trend in peroxide value was noted as heating time increased. A possible explanation for such a trend may be the accumulation of oxidation products in the tallow. Higher concentrations of these products would facilitate their spontaneous reaction wdth newly formed hydroperoxides. Peroxide values for both heating trials may be indicative of both hydroperoxides formed as the 50 N-op x m onm x mmvquucoo - move: r «.me -- can. --- m.~ -- o.o¢ -- com -- o_P -- ea. -- 5., -- m.em mew o.o m.mm mm. om. a.~ m.~ a.c¢ a.mm omN -- «.mo -- as. -- e.~ -- ¢.om mNN e.p¢ F.om «.5. me. o." 4.” m..¢ m.mm oo~ -- m.cm --- 4e. -- M.” -- e.em me. n.“ m.me as. .5. ¢.~ a.~ ¢.~e 5.5m on. -- ~.ce -- he. -- o.m --- o.am mNP c.m c.5m _e. on. m.m m.m N." m.oe co. --- ~.¢m -- me. -- M." -- e..¢ me ¢.a m.~m 44o. no. m.m w.~ o.¢¢ G.F¢ om -- ~.cm -- he. -- e.m -- a.~¢ mu m.e~ o.m~ we. as. c.c o.c m.¢e ”.44 a .ueou .»=H .oeou .o=H .peou .8=H .8eou .o=~ Au¥\aoes Aoe=o=. repeocoo_> Ame e.o< sauce made o=Pa> oeexoeoa o=Pa> deduce weaved: to «see zapped coped: >pucouuweeou=r can xpmzozcvuccu ac mopamegouuogezo pacemaga was peevsogo .m opauh 51 primary oxidation product of unsaturated fatty acyl groups and peroxyl functions formed by the dismutation of such hydroperoxides. Iodine values for the two heating trials clearly indicate that the loss of unsaturation is greater for a given time interval in the intermittently heated tallow (Table 3). The comparative loss of unsaturation in continuous and intermittently heated tallow is represented graphically in Figure 2. Intermittently heated tallow exhibited a decrease in iodine value from 44.3 to 34.8 after 275 hours of heating, representing a decrease of 24.1%. Continuously heated tallow showed a decrease of 10.1%, dropping from 44.5 to 40.0 after 300 hours of heating. This difference in the rate of decrease of iodine value can be attributed to the action of autoxidative reactions in intermittently heated tallow at ambient temperatures. Concommitant with the loss of unsaturation, viscosity of the tallow samples increased (Table 3). This phenomenon was also noted by Perkins (1967). Intermittently heated tallow exhibited a greater increase in viscosity than did continuously heated tallow over a similar period of time. A comparative graphical representation of viscosity changes for both heating trials is shown in Figure 3. Viscosity levels in heated fats and oils are indicative of the amounts of dimeric and polymeric compounds in the lipid moiety (Smouse, 1975). As seen from the data, intermittent heating appears to be more conducive to the formation of these compounds. From the previous discussion, products of autoxidation at ambient temperature, and recurrent localized heating facilitate polymerization reactions in 52 Figure 2. Effect of heating time on the iodine value of tallow heated intennittently (B) and continuously (C). IODINE VALUE 34 53 100 200 300 HEATING TIME (HOURS) AT 180°C 54 .Auv xpmzoacwucou use Amv xpucopumseoucw venom; zappeu do unencumw> ecu :0 were mcwuem; we uumewu .m «gamma. 55 . 0.2: .2 3.505 as: 62:3: . L p b — CON cup our map 00w an an L1 ma 10w -0« . ion 10‘ ran (OO 105 -OO 10¢ r00— 10" ramp (,m :: ems/6 x seloauwao) ulsoosm 56 an intermittently heated fat system. Analysis of iodine values and viscosity data for the two heating trials possibly indicates an inherent difference in the type of reactions occurring in each heating system. Tallow heated intermittently for 125 hours exhibited a 4.5 iodine value unit drop, with a corresponding viscosity increase of 12.2 centipoise x g/cm 3 x 10'2. Tallow heated continuously for 300 hours also exhibited a 4.5 unit dr0p in iodine value, but had a corresponding viscosity increase of 38.0 centipoise x g/cm 3 x 10 ‘2. These data may indicate that, although the loss of unsaturation per unit time is greater in intermittently heated tallow, the reaction involving those double bonds are different in the two heating systems. A greater proportion of the unsaturated linkages lost via continuous heating appeared to be involved in dimerization or polymerization reactions. An equivalent loss of unsaturated bonds via intermittent heating produced approximately one-third the increase in viscosity noted in continuously heated tallow. This may be attributed to the autoxidative reactions encountered in intermittently heated tallow at ambient temperatures. Therefore, the cause for this disparity may be due to the differences in the types of reactions occurring in the two heating systems. Loss of unsaturation in intermittently heated tallow occurs via thermal oxidation and polymerization reactions at elevated temperatures, and by autoxidation reactions at ambient temperatures (Perkins and Van Akkeren, 1965; Perkins, 1967). Continuous heating of tallow allows only thermal oxidative and polymerization reactions, as autoxidation reactions are incapable of occurring under constant frying temperature conditions. 57 Free fatty acid values for the intermittent and continuously heated tallow samples increased with heating time (Table 3). In both trials, the free fatty acid control remained below 1%, which correlated well with the findings of Bates (1952). It was noted that with prolonged heating, the tallow samples became increasingly darker which interfered with accurate endpoint determination by phenolphthalein. Although the recorded values are indicative of the amount of hydrolysis occurring in the triglyceride moiety, free fatty acids may become involved in various decomposition reactions (Chalmers, 1951) or polymerization reactions as indicated by the many studies of fatty acid polymers reviewed by Artman (1969). Short chain fatty acids, liberated via hydrolysis or homolytic cleavage of hydroperoxides, may volatilize at frying temperatures (Chang et al., 1978). Color data for the heated tallow samples are found in Table 4. Color changes were similar for the continuous and intermittently heated tallow samples. It was observed that the tallow which was heated intermittently changed color at a greater rate than the continuously heated samples.Artman (1969) reported that the browning phenemonon accompanies the degradation of a heated fat. Mukai et al.(l965), as cited by Artman (l969),attributed this darkening to the presence of o, a, and o,o unsaturated carbonyl compounds formed via scissioning of hydroperoxides. Also implicated as color contributors in heated fats were highly polar dimeric compounds formed in polymerization reactions of secondary oxidation products (Perkins and Kummerow 1964; Frankel et al., 1960). Table 4. 58 Hunter calorimeter data of continuously and intermittently heated Time of Heating __1. bL (hours) Int. Cent. 'IFE.'-_’7fififii' Int. ‘COnt. 0 +42.7 +42.9 -6.1 -6.0 +13.2 +10.9 25 +40.7 -- -3.9 -- +10.8 -- 50 +40.0 +38.4 -4.2 -3.7 +13.l +14.7 75 +38.9 -- -4.2 -- +16.3 -- 100 +37.9 +35.0 -4.5 -2.2 +18.7 +19.3 125 +36.6 -- -3.5 -- +20.5 -- 150 +34.6 +30.2 -l.3 +3.7 +21.3 +19.2 175 +30.6 -- +4.8 -- +19.7 -- 200 +30.8 +26.8 +4.9 +8.3 +20.1 +17.6 225 +27.6 -- +9.5 -- +18.0 -- 250 +24.9 +23.0 +12.9 +12.8 +16.3 +14.8 275 +18.9 -- +17.2 -- +12.l -- 300 -- +16.l -- +17.l -- +10.0 59 Thompson et al. (1967) reported poor resolution of fatty acid methyl esters derived from heated fats and oils when analyzed by gas chromatography. Hydroxy, keto, dibasic and polymeric acids which form during heating are retained in the column. Further cis-trans isomerization as well as conjugation of polymeric fatty acids have been implicated as sources of poor resolution. Similar results were observed in the heated tallow samples. Analysis by gas-liquid chromatography (GLC) became difficult after only 25 hours of heating in the intermittent system and after 50 hours in the continuous system. Despite the poor resolution, a decrease in unsaturated fatty acids in the sample was noted with prolonged heating (Table 5). This substantiates the literature on the subject as well as the iodine value data obtained in this phase of the study. B. Effects of heating on the stability of cholesterol Concurrent with the oxidatibn of heated tallow triglycerides, the oxidative state of cholesterol present in tallow also changes. Cholesterol levels in tallow have been reported to be 141 mg/kg by Punwar and Derse (1978). Cholesterol under conducive conditions, such as high temperatures and exposure to oxygen, may oxidize to form angiotoxic (Imai et al., 1976; Peng et al., 1978) or carcinogenic compounds (Bischoff, 1969; Smith and Kulig, 1975). Separation of cholesterol and any oxidative derivatives was achieved using the saponification technique described by Itoh et al. (1973). Both continuously and intermittently heated tallow samples and an unheated tallow control were saponified. Analysis of the tallow non-saponifiables was performed using thin-layer chromatography (TLC) and gas-liquid chromatography (GLC). 60 oo.p nu I: a.~m p.59 $.m e.m~ m.~ m.¢ amp mm.o II n: o.am o.np o.op o.- m.~ ¢.¢ ocp cm.c o.~ o.mm m.wp m.P~ ¢.o~ m.~ e.¢ om hm.o -- 5., a..¢ m.m_ m.m o.m~ m.. e.m o . mesa: maozcwucou mo.— nu 1' m.sm ¢.~— p.¢ ¢.a~ m.p s.v amp mc.c u: in m.om N.NF N.m m.- ¢.N ¢.¢ co— om.o as e.~ c.om ~.wp m.o~ m.m~ o.m m.¢ om nw.c in N.F s.~¢ m.mp w.m . o.m~ m.p o.¢ o Auk—.2: =\m nuwp «amp pump cuwp Pump sump u¢p H¢P acmuquLoacu zoppwu cacao; armsocpuccu can xpucoaavsgoucv ea momxpmcm tron xuuum .m «Fae» 61 Table 6. Oxidative products of cholesterol extracted from heated tallow fractions - TLC analysis Rf Color Identity 1.00 I --- Solvent front 0.82 Beige/brown 3,5-Cholestadien-7—onea 0.68 Magenta Cholesterola 0.60 Yellow Unidentifiedb 0.42 Yellow Unidentified 0.28 Blue S-Cholesten-3B,78-diol 0.23 Blue 5-Cholesten-38.7 a-diol 3 compounds present in original taTTOw. b Additional compound present in the 75 hour sample only. TLC analysis of the tallow non-saponifiables indicated that cholesterol subjected to frying conditions is prone to oxidative changes (Table 6). Unheated tallow was found to contain cholesterol with only a trace amount of 3,5-cholestadiene-7-one, indicating that the processes employed in refining tallow did not seriously alter its oxidative state. Analysis of the intermittent and continuously heated tallow non-saponifiable fractions indicated cholesterol underwent oxidation, and that intermittent heating caused a more rapid change in cholesterol stability. This can be attributed to the higher concentration of free radicals in the intermittently heated system. The rate of cholesterol loss and concommitant appearance of oxidation products was greater in the intermittently heated samples as indicated by TLC analysis. It should be noted that similar compound formation and loss in cholesterol was observed in the continuously heated tallow, but at a 62 much slower rate. Since the types of products formed from cholesterol in the heated tallow systems were similar, a representative sample was utilized to investigate the identity of the cholesterol derivatives. The 75-hour intermittently heated fraction was chosen. One dimensional TLC analysis of the intermittently heated tallow non-saponifiable fractions indicated six distinct Spots after 75 hours of heating. However, it was observed that one compound present in the 75 hour sample was not apparent after 150 hours of heating, probably due to decomposition by further oxidation. Also noted was the concommittant decrease in the size of the cholesterol spot with increasing intensity of the oxidation products. The characteristics color and Rf values obtained indicated the presence of 5-cholesten-3B , 7a-diol, 5-cholesten-3B,7B -diol, 5-cholesten-3B -ol, and 3,5- cholestadiene-7-one. Noteworthy was the impurity of the standard cholesterol oxides purchased for this study. The impure state of these compounds hindered both thin-layer and gas-liquid chromatographic analysis of the heated tallow extracts. In order to further establish the identity of the spots, two; dimensional TLC was performed (Table 7). The Rf values for the standards used corresponded to six of the seven spots obtained in the two dimensional analysis. Color development was also compatible with the standard cholesterol oxidation products. Compounds identified included S-cholesten-BB, 7a-diol 5-cholesten-3B,7B -diol, S-cholesten-3B -ol and 3,5—cholestadiene-7-one. Cholestan-3B,Sa,68 -triol was also tentatively identified as the brown spot close to the origin. The low Rf yellow spot may be cholestan-50,6-epoxy -38-ol or 5-cholesten-3B-ol -7—one, but impurities in standards prevented positive identification. 63 Table 7. Two dimensional TLC analyses of oxidation products of cholesterol in tallow heated intermittently for 75 hours and standard compounds. Compound Rf value Color Solvent la Solvent 2h ‘%%}%%%t front 1.00 1.00 3,5-cholestadiene-7-one 0.85 0.72 brown 5-cholesten-3 8 -ol 0.71 0.52 magenta unidentified 0.59 0.52 yellow unidentified 0.47 0.25 yellow 5-cholesten-3 B ,7 B -diol 0.29 0.21 -bl ue 5 cholesten-3 8 ,7a -diol 0.22 0.ll blue cholesten-3 3 , 50. , 6 e -triol 0.05 0.03 brown Standards 3,5—cholestadiene-7-one 0.85 0.72 brown 5-cholesten-3 5 -ol 0.72 0. 53 magenta ‘cholesterol-a-oxide 0.53 0.25 brown/yellow 5-cholesten-38 ,7B-diol 0.30 0.l9 blue 5-cholesten-3 3,7a-diol 0.24 0.ll blue cholesten-3 3, 5a ,63 -triol 0.06 0.04 brown a ethyl acetate - heptane (lzl, v/v) b ethyl acetate - benzene (2:l, v/v) 64 GLC utilizing both capillary and packed columns was employed to analyze the sterol extracts from the heated tallow samples. In both heating trials, an increase in the number of peaks was noted with extended heating. The evolution of peaks occurred at a faster rate in the intermittently heated tallow compared to those samples heated continuously. Chromatograms obtained in capillary GLC analysis for intermittently and continuously heated samples are shown in Figures 4 and 5. Standard compounds were also injected as a means of identification, but results were not conclusive due to the impurity of the standards. Capillary column GLC indicated a greater number of peaks than did packed column GLC. In order to verify the identity of the cholesterol oxidation products formed in heated tallow, compounds were isolated and concentrated using preparative TLC. The samples obtained were subjected to capillary GLC and then to mass spectrometry. The identities of 3,5- cholestadiene ~7-0ne, 5-cholesten-3 B-ol, 5 cholestan-3 , ‘b-diol and 5- cholesten-33, 7 e-diol were tentatively established by these procedures. The mass spectra data obtained for the two unidentified yellow spots were inaccurate due to a variety of interferring compounds, thus no definite identification was possible. 65 Figure 4. Gas chromatograms of tallow non-saponifiable fractions heated intermittently for A) 0 hours, B) 50 hours, C) l00 hours and D) 150 hours. 67 Figure 5. Gas chromatograms of tallow non-saponifiable fractions heated continuously for A) 50 hours, B) 100 hours and C) 150 hours. 68 in l 1 O 5 M 1 10 1 J 15 20 RETENTION TIME (MINUTES) RA. 69 As stated in the literature review, cholesterol as an unsaturated lipid is prone to oxidation. Clearly, the results of this phase of the study indicate that frying conditions, consisting of elevated temperatures in the presence of oxygen, facilitate oxidative changes in cholesterol. Cholesterol oxidation products identified from heated tallow are similar to those reported from aged cholesterol (Horvath, l966; Smith et al., 1967) or intermittently oxidized crystalline cholesterol (Fioriti and Sims, 1967). Cholesterol in tallow heated intermittently or continously follows an oxidation mechanism similar to that given in Figure 1. Attack of a cholesterol molecule by oxygen yields a peroxy radical at the C7 position. It is this reaction which causes the difference in rate of cholesterol oxidation in the two heating trials. The intermittently heated system has a greater oxygen concentration due to incorporation of oxygen via autoxidation at ambient temperatures. Furthermore, the subsequent number of free radicals which may effectively abstract the C7 hydrogen is greater in this system. The abstraction of this hydrogen atom by secondary autoxidation products would greatly facilitate the direct addition of oxygen. Thus, both oxygen and ionizing free radical concentrations are greater in the intermittently heated tallow. Continuously heated tallow incorporates oxygen only by surface entrapment, lowering both the oxygen content and the subsequent radical concentration in the tallow moiety. With the formation of the cholesterol peroxy radical, abstraction of an allylic hydrogen from an unsaturated fatty acyl group may readily occur. The resulting hydroperoxides spontaneously decompose via formal 70 reduction to the epimeric 33 , 7-diols evident in the heated tallow fractions. Similarly decomposition of the 38- 7-hydroperoxides of ‘ cholesterol may readily undergo dehydration reactions due to the high temperature conditions in the fryer. 5~cholesten-3 B-ol-7-one was positively identified by TLC results. The final dehydration product, 3,5-cholestadiene-7-one, was positively identified by two dimensional TLC and GC mass spectrometry. Thus the existence of 5—cholesten-3 s—ol-7-one in heated tallow is suspected to be transient at such elevated temperatures. ‘ Smith (1980) stated that reaction of cholesterol-7-hydroperoxide with an allylic hydrogen or hyderoperoxide may cause the formation of the epimeric cholestan~5, 6a -epoxy-3B —ol and cholestan-5,68 -epoxy -38-ol. Although these compounds were not positively identified in the heated tallow fractions, their presence is suggested by the identification of their hydration product, cholestan-3B , 5o ,GB-triol, in the intermittently heated tallow. Water, formed as a product of various oxidation reactions, may be present in heated tallow systems. Thus, the formation ofo or B-oxides of cholesterol in heated tallow systems may be followed by spontaneous hydration, yielding the triol species. With regards to the analytical procedures employed in this study, it appears that two dimensional TLC analysis is sufficient to positively identify cholesterol oxidation products. Smith (personal communication) has indicated that Rf values and color development after acid treatment adequately confirm the identities of sterols when compared to standard references. Various anomalies have been reported by researchers concerning alternate techniques of analysis of sterols. 71 Fioriti and Sims (1967) noted the behavior of cholesterol autoxidation compounds during GLC analysis. 5—Cholesten-38 ,7a -diol and 5-cholesten-38, 7B -diols were found to undergo an on-column decompositon, with the decomposition products stable at operating temperatures. Claude (1966) observed that the 3B , 7-diols decomposed to products having retention times similar to 3,5-cholestadiene-7-one, indicating the formation of a diene-like species. Similarly, Van Lier and Smith (1968) noted the conversion of 5-cholesten-3B -ol-7-one to 3,5- cholestadiene via on-column dehydration. The same reseachers also stated that the presence of the 3B, 7-diols cannot be confirmed due to the previously noted on-column decomposition. In conclusion, it was stated that no single system or derivatization procedure adequately separates important compounds formed in cholesterol oxidation. Sheppard and Shin (l980) identified the inherent impurities of standard cholesterol oxidation products as a major deterrent to elucidation of the cholesterol oxidative mechanism. Impurities in the standards were encountered in this study which seriously hindered GLC and GC-MS investigations. 72 Phase II. The Effects of Intermittent Frying on the Oxidative Stability of Tallow. In the first phase of this study, it was established that intermittent heating of tallow resulted in the rapid degradation of the. fat and oxidation of cholesterol. In order to replicate commercial practices more closely, a study involving the production of french fries by frying in intermittently heated tallow was conducted. A. Effects of Frying in Tallow The results of the analyses performed on the tallow samples indicate that frying had a suppressive effect on deterioration. Physical and chemical data showed that the action of frying in intermittently heated tallow resulted in a less rapid deterioration than tallow heated intermittently without the introduction of fried substrate. The tallow used for frying did however deteriorate faster than the continuously heated tallow studied in Phase I. The introduction of fried products to an intermittently heated system results in extension of the fat life due to the constant absorption of the frying medium by the substrate and the subsequent addition of fresh fat or oil to the fryer (Melnick, l957b; Artman, l969). Absorption of fat by the subtrate also results in the absorption of degradation products (Hussain and Norton, 1976). Add-back of fat dilutes the degradation products, as well as providing a fresh matrix for frying. Greater frying loads results in more add-back of fat to the fryer which results in a longer frying life (Melnick, l957b). In addition, steam generated from water in the food substrate may carry volatile degradation products from the fat (Chang et al., l978), as well as occlude oxygen from the fryer surface, deterring oxidative reactiOns (Melnick et al., 1958). 73 Peroxide values for the tallow, presented in Table 8, reached a maximum of 3.8 in the ZOO-hour study. This low value is the result of factors previously discussed in Phase I of this study. The introduction of french fried potatoes into the tallow or the replacement of fresh tallow into the fryer did not seen to influence the rate of peroxide formation. Iodine values of the tallow decreased steadily with extended exposure to frying temperature (Table 8). The tallow showed a dr0p in iodine value from 44.2 initially to 40.9 after l50 hours of heating representing a 7.5% decrease. Tallow samples heated intermittently and continuously exhibited 10.5% and 4.6% decrease, respectively (Table 3). Graphical representation of these data is found in Figure 6. Simdlarly, viscosity levels of the utilized tallow increased with prolonged heating (Table 8). This increase was noted as the iodine values for the samples decreased. A comparison of the viscosity data from the used intermittently heated tallow and the viscosity data derived from the two heating trials in Phase I is presented in Figure 7. The tallow in which french fry potatoes were processed increased in viscosity from 27.0 to 5l.3 centipoise x g/cm3 x 10 '2 after 200 hours of heating, representing a 24.3 unit increase. Intermittent and (continuously heated tallow exhibited a 3l.l and 14.4 unit increase after 200 hours of heating, respectively (Table 3). Free fatty acid values for the utilized tallow indicated no increase in hydrolytic activity due to the introduction of water into the frying medium via french fry potatoes (Table 8). Values did not differ appreciably from those obtained in the analysis of continuous and intermittently heated tallow to which no product was introduced. 74 Niop msu\m x ochFucou n mp—c: a eesee ea xeee eeeee zeppee « m.Pm ¢~.o p.~ n.mm cow @.m¢ Nu.o o.F o.oe mNP ~.cm P~.o p.~ m.o¢ aomp m.om se.o ~.~ ~.p¢ mup w.mm mm.o m.~ o.~¢ «cop c.pm mo.o ¢.~ ¢.~¢ mm c.o~ mw.o m.m N.m¢ om ¢.w« mw.o ~.~ c.¢¢ mm c.s~ o~.c c.o ~.¢¢ o Aw. .ax\aeev Ameaoz. exu'moum*> vvo< xuuau omen o=Fu> meexogca o=Pa> ocvuon a:_peoz mo meek xespm acrxge av com: zappaa mo mu_umveapuogesu pacesogu new pauvnge .m o—aoe 75 Figure 6. Effect of heating time on the iodine value of tallow heated intermittently B), continuously C), and intermittently with frying D). 76 45 44 . 4a « 42 « 41 4 4o -: 39- IODINE VALUE 38‘ 37‘ 36- 35‘ 34 O 100 200 300 HEATING TIME (HOURS) AT 180°C 77 .Aa meexge new: »_ucmuupsemuce use Au »_m:o::_u:ou .Am xpucmuu_semucw vogue; zap—cu mo xu_moumv> any so we?“ mcwuam; mo Heaven .5 me=m_m 78 ab“ P Gnu Decor .2 Amp—=05 NEE. 02:39.. OOH Our 00' On 0, b — lliili‘hilll Illillil .IIIIi icw rON 00 i0? 106 row was 100 iOOw iOpp r¢up :00 . (z-OI- "3419/6 X SSIOdILNaO) ALISOOSIA 79 Although these results indicate that hydrolytic activity did not increase, loss of free fatty acids by steam distillation or absorption by the substrate may compensate for the low free fatty acid values (Chang et al., 1978; Pokorny, 1980). Color data for the utilized tallow are found in Table 9. These data indicate a darkening of the tallow as heating time progressed, as previously noted for heated tallow used in Phase I of this study. Intense foaming occurred in the tallow medium after l25 hours of intermittent heating. This phenomenon can be attributed to the buildup of polymeric and dimeric compounds in the tallow matrix as evidenced by increases in viscosity levels (Morton, 1977). Smouse (l975) has reported that various vegetable oils foamed after similar heating. This was attributed to polymeric substances as well as free fatty acids in the heated oil system. Analysis of the non-saponifiable fractions from the tallow samples indicated that oxidation of cholesterol occurred during the frying procedure (Table 10). TLC analyses confirmed the presence of 5- cholesten-3B , 7o -diol, 5-cholesten-3B,73-diol, 5-cholesten-33-ol and 3,5-cholestadiene-7-one in the tallow samples after 25 hours of heating. An unidentified yellow spot was also observed with a similar Rf value to the unknown compound noted previously (Table 6). An additional magenta spot was isolated which could not be identified by reference standards. All oxidation products increased with prolonged heating at the expense of cholesterol. 80 Table 9. Hunter colorimeter data for tallow utilized for frying Time (Hours) L aL bL 0 +42.l -4.l +2.2 25 +37.0 -3.5 +5.6 50 +33.2 -0.8 +5.2 75 +28.7 +3.9 +l.7 100 +24.l +8.l -4.7 125 +Zl.l +lO.6 -9.6 150 +l9.4 +ll.8 -12.6 175 +l7.4 +l2.7 -l7.3 200 '+l5.7 4+l3.3 -2l.4 Table 10. TLC analysis of the non-saponifiable fractions of tallow used in frying Rf Value Color Intensity with Time Identity l.00 ~--- -- solvent front 0.98 brown increase 3,5-cholestadiene-7-one 0.95 magenta no change unidentified 0. 87 magenta decrease 5-chol esten-3 B—ol 0.68 yellow increase unidentified 0.47 blue increase 5-cbolesten-3 8,3 -diol 0.39 blue increase 5-cholesten-38, 7o-diol 81 GLC results of the non-saponifiable fractions from the tallow indicate that the overall sterol profile remained somewhat constant for the duration of the study (Figure 8). An increase in the peak eluting immediately after cholesterol was noted. This peak also increased with extended heating in both the intermittent and continously heated samples (Figure 4 and 5). A series of small peaks eluting after the cholesterol peak may be due to plant sterols which migrated from the french fried potatoes to the tallow during frying. The chromatogram of the non-saponifiable fraction of unfried french fry potatoes lipids is found in Figure 9 and will be discussed later in the text. 8. Analysis of French Fried Potatoes The results of the fat absorption of french fry potatoes fried in tallow are found in Table ll. The fat content of the french fried potatoes ranged from 19.78 to 24.48%, which was appreciably higher than fat contents reported for commercially prepared french fried potatoes by Slaver et al. (l980). The noted higher fat content may be attributed to the larger size and greater surface area of the crinkle cut potatoes used in this study. TLC analysis of the non-saponifiable fractions of the fat absorbed by french fried potatoes showed that the oxidation products formed in tallow were absorbed by the french fried potatoes as well (Table 12). Similar to the sterol profiles for the tallow samples, all oxidation products present increased with prolonged heating at the expense of cholesterol. 82 Figure 8. Gas chromatograms of tallow non-saponifiable fractions from tallow samples intermittently heated for frying. (A) 50 hours, (B) lOO hours and (C) 150 hours 83 a c r - M a W I I I I J I I I I J 0 5 ‘IO 15 20 0 5 10 15 20 “I l A r “W l l I l J O 5 10 15 20 RETENTION TIME (MINUTES) 84 Table 11. Absorption of fat by french fry potatoes fried in tallow '— Sample Time Range of Grams of Fat in . %Fat Number Frying (Hours) 400 G Sample Control --- 10.1 2.5 l 0-25 79.1 19.8 2 25-50 87.5 21.9 3 50-75 97.9 24.5 4 75-100 81.1 *20.3 5 100-125 86.7 21.7 6 125-150 81.6 *20.4 7 150-175 88.8 22.2 8 175-200 90.2 22.6 *Fresh tallow added back to fryer Table 12, TLC analysis of the non-saponifiable fraction of fat extracted from tallow-fried french fried potatoes Rf Value Color Intensity with time Identity 1.00 -- -- solvent front .99 beige/brown increase 3,5-cholestadiene-7-one .96 magenta no change unidentified .87 magenta decrease 5-cholesten-3B-ol .69 yellow increase unidentified .48 blue increase 5-cholesten-38, 7B-diol .41 blue increase 5-cholesten-3B, 7a—diol .04 brown decrease cholestan-3B, 50,68-triol 85 Figure 9. Gas chromatograms of lipid non-saponifiables extracted from french fried potatoes fried in tallow heated intermittently for (B) 50 hours, (C) 100 hours, (D) 150 hours and (A) unfried control. 86 LkBr MR0 m A , he I ' l _I I l l I J l J O 5 1O 15 20 O 5 1O 15 20 RETENTION TIME RETENTION TIME (MINUTES) (MINUTES) 87 5-Cholesten-3l; 7cediol, 5-cholesten-3B, 7B-diol, 5-cholesten-38 -01 and 3,5-cholestadiene-7-one were identified in the french fry potatoes lipid extracts. Cholestan-3B, 54, 6B -triol was also detected in trace amounts. This compound, as previously described, is the hydration product of cholestan-5,6a -epoxy-3B-ol. The presence of this compound in the french fry potato extract is understandable due to the high concentration of water in the potato matrix. Chromatograms from the GLC analysis of the french fry potatoes lipid extracts are shown in Figure 9. As seen from these chromatograms, the lipid extract from unfried french fry potatoes contained a number of compounds which could possibly be plant sterols as well as cholesterol. The french fried potatoes utilized in this study were treated with vegetable shortening containing soybean, palm, or cottonseed oil, and beef fat and may adequately account for the sterol profile of the uncooked french fry lipid extract. As a confirmation of vegetable oils in the french fry potato fatty acid analysis was performed on the lipid extracts (Figure 10). The results of the analyses indicate a high concentration of unsaturated fatty acids compared to similar data for tallow (Table 13). The migration of plant sterols from the french fried potatoes into the heated tallow moiety may account for some of the small peaks observed in the GLC sterol profiles for the tallow and lipid extracts from the french fried potatoes. The chromatograms obtained for the lipid extracts for french fried potatoes exhibited sterol profiles similar to those for the tallow samples. Figure 10. Gas chromatogram of fatty acid methyl esters from a lipid extract of unfried french fry potatoes. 89 9.30 «flavor taper 95.0 OuOPUI 930 90 Table 13. Fatty acid analysis of tallow and lipid extracts from raw french fry potatoes. Fatty Acid C14 C14:l C16 016:1 C18 C18:1 C18:2 C18:3 SflJ Tallow 4.0 1.3 26.6 8.8 15.9 41.7 1.7 --- 0.87 French fried 1.1 0.2 15.1 1.5 8.6 43.6 26.1 2.7 0.33 potato extract GLC analysis indicated that the concentration of cholesterol oxidation products in the lipid extracts from french fried potatoes was approximately four times greater than the corresponding tallow samples. A dilution factor of four was required for the french fried potato lipid extracts in order to obtain peak sizes similar to the undiluted tallow extracts. This concentration of the cholesterol and its oxidation products in the french fries is in agreement with the findings of Hussein and Morton (1976) who reported that oil absorbed in fried foods is more oxidized than the frying medium itself. Similarly, Pokorny (1980) indicated that preferential absorption of frying fat decomposition products may occur due to polar interaction with the substrate. The increased polarity of cholesterol and its oxidation products relative to the triglyceride moiety may account for this partition effect. 91 Phase III. The Occurrence of Cholesterol Oxidation Products in Commercially Produced French Fried Potatoes °Results of Phase I and II of this study indicate that intermittent heating facilitates the decomposition of tallow used as a frying medium and the cholesterol contained therein. It was also found that these cholesterol oxidation products are preferentially absorbed by french fried potatoes fried in the heated tallow. To evaluate the implications of this study, commercial french fried potato samples processed by frying in intermittently heated tallow, were obtained and analyzed for the presence of cholesterol oxidation products. Chromatographic analysis of a non-saponifiable fraction of frying medium samples from a fast-food franchise indicated that beef tallow was utilized as the frying fat (Figures 11 and 12). Thus, french fried potatoes samples from two local fast-food franchises were collected over a two week period and analyzed for cholesterol oxidation products. Fat absorption data for the two sets of french fried potatoes samples are found in Table 14. Fat content of the french fried potatoes ranged from 16.4 to 20.9%. These values are higher than those obtained by Slover et al., (1980) who reported levels of 11.94 to 13.76% fat in commercial french fried potatoes. This fat content was also substantially lower than those recorded for french fried potatoes analyzed in Phase II of this study (Table 11). The difference may be attributed to the larger size and surface area of the crinkle cut potatoes used in the laboratory study. The influence of size/surface area on the absorbed fat content was noted in the works of Strock et al., (1966) who reported 7% fat in 12 mm fried sliced potatoes, and Artman (1969) who found 1 mm slices to contain 30-40% absorbed fat. 92 Figure 11. Gas chromatograms of the non-saponifiable fraction from new fast-food franchise frying medium. 93 I l ' I l I O 5 10 15 2O RETENTION TIME (MINUTES) 94 Figure 12. Gas chromatogram of the non-saponifiable fraction from discarded fast-food franchise frying medium. 95 M l I I s 10 15 RETENTION TIME (MINUTES) 20 96 Table 14. Fat absorption data of commercially fried french fry potatoes Sample Sequence Grams of Fat in 150 9 Sample . % Fat A B ‘ A B A B l 1 24.6 25.7 16.4 17.1 2 2 27.5 4 26.3 18.3 17.5 3 3 31.4 25.2 20.9 16.8 4 4 25.1 29.1 16.7 19.4 5 5 27.9 30.6 18.6 20.4 6 6 25.0 25.6 16.7 17 1 The extracted fat from a fast-food franchise french fried potatoes exhibited color changes over the two week sampling period (Figure 13). It is a evident by the color differences that the lighter sample represents a new batch of the frying medium. It is assumed that the lightest sample indicates the use of a new medium and is therefore indicative of heating time. Variations in color degradation may be attributed to increased use of the medium resulting in the need for more frequent replenishment with fresh fat. TLC analysis of the non-saponifiable fraction of the lipids extracted from the fast-food franchise french fried potato samples indicated the presence of 5-cholesten-3B, 7B -diol, 5-cholesten-38 , 7B -diol, 5-cholesten-38 -01, and 3,5-cholestadiene-7-one in the french fried potatoes (Table 15). No cholestan-3B, So , 6B-triol was detected. The cholesterol spot was noted to decrease in color intensity as the lipid extracts became darker. The unidentified yellow spot (Rf =0.63) encountered in all previous heated tallow extracts was also present in both sets of french fry samples. An additional unidentified brown spot (Rf =0.88) was also detected in the non-saponifiable fractions of the two series of samples. A yellow spot R f = 0.92, was observed in only one of the series of samples. The presence of these 97 Figure 13. Lipid extracts from french fried potatoes obtained from (A) an East Lansing fast-food restaurant and (B) and Okemos area fast-food restaurant. 99 Table 15. TLC analysis of the non-saponifiable fraction of fat extracted from commerCial french fry potatoes Rf Value Color Intensity A B with time Identity 1.00 -- -- solvent front 0.98 0.99 brown increase 3,5-cholestadiene-7-one 0.92 -- yellow decrease unidentified 0.88 0.88 brown no change unidentified 0.85 0.84 magenta decrease S-cholesten-BB-ol 0.64 0.63 yellow increase unidentified 0.47 0.46 blue increase 5-cholesten-3B, 7B-diol 0.40 0.39 blue increase cholestan-38, 7a-diol compounds may be attributed to many factors. Hydrogenation of tallow prior to use may form sterol derivatives. The introduction of fish or fruit products into the tallow for frying may result in the migration of sterols from these products into the tallow moiety. These may eventually be absorbed by french fried potatoes fried in the same oil. The results of GLC analyses of non-saponifiable fractions from lipid extracts of a fast-food franchise french fried potato samples are found in Figure 14. The chromatograms, representing the darkest lipid sample, show a sterol profile similar to those obtained for lipid extracts from the french fried potatoes produced in the laboratory (Figure 9). A dilution factor similar to that used in Phase II of this study was also needed to facilitate GLC analysis. 100 Figure 14. Gas chromatograms of lipid non-saponifiables extracted from area fast-food franchises french fried potatoes, (A) East Lansing area and (B) Okemos area. 1 \I J I I 5 1O 15 RETENTION TIME (MINUTES) 101 I 20 It I l I I I O 5 10 15 20 RETENTION TIME (MINUTES), 102 The result of this study indicates that cholesterol oxidation products are formed in fried foods produced under commercial practices when tallow is used as the frying medium. SUMMARY AND CONCLUSIONS The oxidative stability of the triglycerides and cholesterol in edible tallow employed as a deep-fat frying medium was investigated in a three-phase study. Intermittent heating was found to be more detrimental than continuous heating to tallow in regard to frying stability. Chemical and physical analyses including iodine value, viscosity, peroxide value, free fatty acid, fatty acid analysis, and color indicated that the accumulation of products from degradation reactions occurred more rapidly in intermittently heated tallow. These studies also showed that cholesterol present in tallow readily oxidizes under conditions utilized in frying operations. TLC and GLC analyses indicated that intermittent heating promoted more rapid changes in the oxidative state of cholesterol. Various oxides including 5-cholesten-3i3, 7cx-diol, 5-cholesten-3B , 7 B-diol, 5-cholesten-33 -ol, 3,5-cholestadiene-7-one and cholestan-3 B , 50 ,6 B-triol were identified in the heated tallow samples by two dimensional TLC. In the second phase of the study, the effect of frying on the oxidative stability of intermittently heated tallow triglycerides and cholesterol was investigated. The act of frying french fry potatoes in intermittently heated tallow was observed to have a suppressive effect on the degradation of the tallow. This was attributed to the absorption of deleterious compounds by the french fried potatoes and the dilution of remaining degradation products by the addition of fresh tallow for that depleted by absorption. TLC analysis of tallow and french fry potato samples indicated that cholesterol oxidizes in the system and 103 104 that the formed oxidative derivatives of cholesterol are preferentially absorbed by french fried potatoes. Oxides including 5-cholesten-38, 7a -di01, 5- cholestan-33,7B-diol, 5-cholesten-3B -01, and 3,5-cholestadiene-7-one were identified on both tallow and french fried potato samples. Cholestan-3B, 5a, 63-triol was also identified in trace amounts in french fried potato. In the third phase of the study, commercially prepared french fried potatoes from a franchise utilizing tallow as a frying medium were investigated and were found to contain 5-cholesten-38 , 7o-diol, 5- cholesten-Be , 7a -diol, 5-cholesten-3B -ol and 3,5-cholestadiene-7-one. The conclusions reached as a result of this study are summarized below: 1) Intermittent heating of tallow facilitates a more rapid degradation of tallow triglycerides and oxidation of cholesterol than does continuous heating due to autoxidation at ambient temperatures. 2) A greater proportion of unsaturated linkages lost during heating are consumed in dimerization and polymerization in continuously heated systems than in intermittently heated systems, while the rate of such reactions proceeds much more slowly in the continuously heated system. 3) Oxidation of cholesterol readily occurs at conditions suitable for frying and 5-cholesten-3B,7o-diol, 5-cholesten-3B,7B-diol, and 3,5- cholestadiene-7-one are ubiquitous to heated tallow systems. 4) The introduction of a substrate into heated tallow has a suppressive effect on the degradation of the fat due to absorption by the substrate and dilution by fresh tallow of the degradation products which accumulate with prolonged heating. 105 5) Cholesterol and cholesterol oxidation products are apparently preferentially absorbed by french fry potatoes fried in the heated tallow moiety. The sterols are approximately four times more concentrated in the french fry potato lipid extract than the heated frying medium. 6) The identified cholesterol oxidation products present in heated tallow systems have not been implicated as having atherosclerotic or carcinogenic properties, although they have been isolated from the lipid fraction of human atheromata. 7) Trace amount of cholestan-38 , 50, 68 -triol were noted in french fried potatoes fried in tallow. This compound has been well documented as being a potent atherosclerotic and angiotoxic agent. Its presence also indicates the transient existence of its precursor, cholestan-5,6a-epoxy-3 e-ol, a known carcinogen, in heated tallow systems. 8) Commercially processed french fried potatoes which are fried in tallow such as those distributed through fast-food franchises, contain oxidation products of cholesterol including 5-cholesten-3B, 7a-diol, 5- cholesten-3s,7s -diol and 3,5-cholestadiene-7-one, as well as unoxidized cholesterol. PROPOSALS FOR FURTHER RESEARCH The study of tallow constituent oxidation during frying processes has raised some questions worthy of further investigation. These include: 1. The possible participation of cholesterol or cholesterol oxidation products in polymerization reactions occurring in heated tallow. 2. The identification and isolation of the different dimeric and polymeric species formed in intermittent and continously heated tallow. 3. The effect of unsaturated bond concentration on the oxidative stability of cholesterol in a heated tallow system. 4. In this study, several compounds were isolated which could not be identified by standard references. Application of GC-MS techniques with free sterols or silylated derivatives may allow resolution of these structures. 5. It has been established that cholesterol in tallow migrates to products fried in tallow. Research is needed to determine if cholesterol-containing foods such as chicken or fish release cholesterol to the frying medium which may eventually be reabsorbed by other food products. 6. Extensive work in the area of cholesterol oxidation is needed in foodstuffs stored at ambient temperatures such as dried sausage and cheese products. 106 107 7. An investigation of plant sterols in fried food systems, either naturally occurring in plant oils or via migration from fried vegetable products merits more study. 8. The development of preparative procedures which would allow separation of sterols from lipid matrices and prevent the need for saponification would prove invaluable in future studies. 9. More extensive research in the area of the biological effects of the cholesterol oxidation products formed in all processed foods is needed to assess the true hazard these compounds may pose in the diet. It is important to establish whether cholesterol or its oxidation products are responsible for atherosclerotic, angiotoxic, and carcinogenic effects. BIBLIOGRAPHY Acker, L. and Greve, H. 1963. Uber die photooxydation des cholesterins in eihaltigin Lebensmitteln. Fette Seifen Anstrech. 65:1009. Ansari, G.A.S. and Smith, L.L. 1979. The oxidation of cholesterol by hydroxyl radical. Photochem. Photobiol. 30:147. Artman, N.R. 1969. The chemical and biological pr0perties of heated and oxidized fats. Adv. Lipid Res. 7:245. Bates, R.N. 1952. Quality deep fat-fried foods. Food Eng. 24(12):82. Bergstrom, S. and Samuelsson, B. 1961. The autoxidation of cholesterol- In “Autoxidation and Antioxidants," N.0. Lundberg, Ed., Interscience Publizgers, John Wiley andSons,New York City, N.Y., Vol. 1 pp. 233-2 . Bischoff, F. 1963. Carcinogenesis through cholesterol and derivatives. Prog. Exp. Tumor Res. 3:412. Bischoff, F. 1969. Carcinogenic effects of steroids. Advan. Lipid. Res. 7:165. Black, H.S. and Douglas, B.R. 1972. A model system for the evaluation 3; ggggesterol a-oxide in ultraviolet carcinogenesis. Cancer Res. Black, H.S. and Douglas, B.R. 1973. Formation of a carcinogen of natural origin in the etiology of ultraviolet light-induced carcinogenisis. Cancer Res. 33:2094. Bladon, P. 1958. The superficial oxidation of cholesterol. In "Cholesterol," R.P. Cook, Ed. Academic Press Inc. New York. p.76. Brooks, C.J.N., Harland, N.A. and Steel, G. 1966. Squalene, 26-hydroxy- cholesterol and 7-ketocholesterol in human atheromatous placques. Biochim. BiOphys. Acta. 125:620. Brooks, C.J.N., Steel, 6., Gilbert, J.D. and Harland, w.A. 1971. Lipids of human atheromata. Part 4: Characterization of a new group of polar sterol esters from human athersclerotic placques. Atherosclerosis. 13:223. Chalmers, J.G. 1951. Chemical changes in cottonseed oil on heating to various temperatures. Acta Unio Intern. Contra Cancrum. 7:612. Chan, J.T. and Black, H.S. 1974. Skin carcinogenesis: cholesterol-50, 60.-epoxide hydrase activity in mice skin irradiated with ultraviolet light. Science. 186:1716. 108 109 Chan, J.T. and Black, H.S. 1976. Distribution of cholesterol 5a, 6cbepoxége fgrmation and its metabolism in mouse skin. J. Invest. Derm. :11 . Chang, 5.5. and Kummerow, F.A. 1953. The isolation and characterization of the polymers formed during the autoxidation of ethyl linoleate. J. Amer. Oil Chem. Soc. 30:403. Chen, H.N., Kandutsch, A.A. and Naymouth, C. 1974. Inhibition of cell growth by oxygenated derivatives of cholesterol. Nature. 251:419. Chicoye, E., Powrie, N.D. and Fennema, O. 1968. Photooxidation of cholesterol in spray-dried egg yolk upon irradiation. J. Food Sci. 33:581. Claude, J.R. 1966. Separation et identification de sterols substitves sur les cycles Aet 8 par chromatographie en couches Minees et chromatographie en phase gazeuse. J. Chromatog. 23:267. Crampton, E.N., Common, R.H., Farmer, F.A., Nells, A.E. and Crawford, O. 1953. Studies to determine the nature of the damage to the nutritive value of some vegetable oils from heat treatment. J. Nutr. 49:333. Deuel, H.J. Jr., Greenberg, S.N., Colbert, C.E., Baker, R. and Fisher, N.R. 1951. Toxicology studies on isopropyl and stearyl citrate. Food Res. 16:258. Evans, C.O., McConnell, O.G., Frankel, E.N. and Cowen, J.C. 1965. Cbromatrograpic studies on oxidative and thermal fatty acid dimers. J. Amer. Oil Chem. Soc. 42:764. Farmer, E.N., Bloomfield, G.F., Sundralingam, A. and Sutton, O.A. 1942. The course and mechanism of autoxidation reactions in olefinic and polyolefinic substances including rubber. Trans. Faraday Soc. 38, 348:356. Fioriti, J.A. and Sims, R.J. 1967. Autoxidation products of cholesterol. J. Amer. Oil Chem. Soc. 44:221. Firestone, 0., Horwitz, N., Freidman, L. and Shue, G.M. 1961. Heated fats. 1. Studies of the effects of heating on the chemical nature of cottonseed oil. J. Amer. Oil Chem. Soc. 38:253. Flanagan, V.P., Ferritti, A., Schwartz, D.P. and Ruth, J.M. 1975. Characterization of two steroidal ketones and two isoprenoid alcohols in dairy products. J. Lipid Res. 16:97. Frankel, E.N., Evans, C.O. and Cowan, J.C. 1960. Thermal dimerization of fatty ester hydroperoxides. J. Amer. Oil Chem. Soc. 37:418. Gunstone, F.D. and Hilditch, T.P. 1945. The union of gaseous oxygen with methyl oleate, linoleate and linolenate. J. Chem. Soc. 836. 110 Henderson, A.E. 1956. A histochemdcal and chromatographic study of gorggl and atheromatous human arteries. J. Histochem. Cytochem. :1 . Henderson, A.E. and MacDougall, J.B.D. 1954. A histochemical and chromatographic study of the lipid distribution in human arteries. Biochemical J. 57:xxi. Horvath, C. 1966. Quantitative determination of cholesterol in autoxi- dation mixtures by thin-layer chromatography. J. Chromatog. 22:52. Hussain, 5.5. and Morton, 1.0. 1976. Absorption of food by frying. Proc. 4th. Internat. Cong. Food Sci. and Tech. Madrid, Spain Vol. I. Imai, H., Nerthessen, N.T., Subramanyan, V. LeOuesne, P.N., Solaway, A.H. and Kanesawa, M. 1980. Angiotoxicity of oxygenated sterols and possible precursors. Science 207:651. Imai, H., Nerthessen, N.T., Taylor, C.B. and Lee, K.T. 1976. Angiotoxicity and athersclerosis due to contamdnants of USP-grade cholesterol. Arch. Pathol. Lab. Med. 100:565. Itoh, T., Tamura, T. and Matsumato, T. 1973. Methylsterol composition of 19 vegetable oils. J. Amer. Oil Chem. Soc. 50:300. Kandutsch, A.A., Chen, H.N. and Heiniger, H.J. 1978. Biological activity of some oxygenated sterols. Science. 201:498. Kandutsch, A.A. and Chen, H.N. 1978. Inhibition of cholesterol synthesis by oxygenated sterols. Lipids. 13:704. Kartha, A.R.S. 1960. The iodine value decreasing reaction in fat autoxidation. J. Sci. Research (India) 198, 199-205. Kaunitz, H., Slanetz, C.A., Johnson, R.E., Knight, H.B., Saunders, D.H. and Swern, D. 1956. Biological effects of the polymeric residues isolated from autoxidized fats. J. Amer. Oil Chem. Soc. 33:630. Keane, K.N., Jacobson, G.A. and Krieger, C.H. 1959. Biological and chemical studies on commercial frying oil. J. Nutr. 68:57. Kirby, A.H.M. 1943. Attempts to induce stomach tumors. 1. The effect of cholesterol heated to 300°C. Cancer Res. 3:519. Kirby, A.H.M. 1944. Attempts to induce stomach tumors. III. The effects of a) a residue of cholesterol heated to 300°C., and b) 3,5-cholestadiene. Cancer Res. 4:94. Kulig, H.J. and Smith, L.L. 1973. Sterol metabolism. XXV. Cholesterol oxidation by singlet molecular oxygen. J. Org. Chem. 38:3639. 111 Kummerow, F.A. 1979a. Nutrition inbalance and antiotoxins as dietary risk factors in coronary heart disease. Am. J. Clin. Nutr. 35:58. Kummerow, F.A. 1979b. Possible role of oxidized lipids in atherogenesis. Paper presented at the workshop, Autoxidation Processes in Food and Related Biological Systems. H.S. Army Research and Dev. CoIIIIIand, Natick, Mass. Lane, A., Blickenstaff, D. and Ivy, D.C. 1950. The carcinogenecity of fats "browned" by heating. Cancer 4:1044. Larsen, C.D. and Barrett, M.K. 1944. Administration of 3,5- cholestadiene and dicholesteryl ether to mice and rats. J. Nat. Cancer Inst. 4:587. MacDougall, J.D.B., Biswas, S. and Cook, R.P. 1965. The effect of certain C27 steroids on organ cultures of rabbit aorta. Br. J. Exp. Pathol . 46:549. Melnick, O. l957a. Absence of thermal polymers in potato-chip frying oils. J. Amer. Oil Chem. Soc. 34:351. Melnick, O. l957b. Nutritional quality of frying fats in commercial use. J. Amer. Oil Chem. Soc. 34:578. Melnick D., Luckmann, F.H. and Gooding, C.H. 1958. Composition and control of potato chip frying oils in continuing commercial use. J. Amer. Oil Chem. Soc. 35:271. Micheal, N.R. 1966. Thermal reactions of methyl linoleate. 11. The structure of aromatic C13 methyl esters. Lipids 1:359. Micheal, N.R. 1966. Thermal reactions of methyl linoleate III. Characterization of 013 cyclic esters. Lipids 1:365. Morrison, N.R. and Smith, L.M. 1964. Preparation of fatty acid methyl esters and dimethylacetals from lipids with boron-flouride-methanol. J. Lipid Res. 5:600. Morton, 1.0. 1977. Changes related to different time-temperature combinations. In Physical, Chemical and Biological Chan es in Food Caused b Thermal Processin . ‘H0yem, T; andfiKvale, O.- -ds. ‘AppTTed ScIence PuE. ETD. Essex Eng. pp. 135-151. Nolen, G.A., Alexander, J.C. and Artman, N.R. 1967. Long-term rat feeding study with used frying fats. J. Nutr. 93:337. O'Neill, L.A. 1954. Some recent studies on the autoxidation of drying oils. Chem. Ind. (London). p. 384. Peng, S.K., Taylor, C.B., Tham, P. and Mikkelson, B. 1978. Effect of autoxidation products from USP-Grade cholesterol on aortic smooth muscle cells. An in vitro study. Arch. Pathol. Lab. Med. l02:57. 112 Peng, S.K., Tham, P., Taylor, C.B. and Mikkelson, B. 1979. Cytotoxicity of oxidation derivatives of cholesterol on cultured aortic smooth muscle cells and their effect on cholesterol biosynthesis. Am. J. Clin. Nutr. 32:1033. _ Perkins, E.G. 1960. Nutritional and chemcial changes occurring in heated fats: A review. Food Technol. 14:508. Perkins, E.G. 1967. Formation of non-volatile decomposition products in heated fats and oils. Food Technol. 21:611. Perkins, E.G. and Kummerow, F.A. 1959. The isolation and characterization of the polymers formed during the thermal oxidation of corn oil. J. Amer. Oil Chem. Soc. 36:371. Perkins, E.G. and Van Akkeren, L.A. 1965. Heated fats. IV. Chemical changes in fat subjected to deep fat frying processes: Cottonseed oil. J. Amer. Oil Chem. Soc. 42:782. Pokorny, J. 1966. The autoxidation of some vegetable oils at elevated temperatures. XIII. Effect of temperature and diffusion on the zgiggg ghanges during the oxidation of rapeseed oil. Chem. Abstr. : l . Pokorny, J. 1980. Effect of substrates on changes of fats and oils gaging frying. La Rivista Italiana Delle Sastanze Grasse. LVII Poling, C.E., Warner, N.D., Mone, P.E. and Rice, E.E. 1960. The nutritional value of fats after use in commercial deep-fat frying. J. Nutr. 72:109. Poling, C.E., Narner, N.D., Mone, P.E. and Rice, E.E. 1962. The influence of temperature, heating time and aeration upon the nutritive value of fats. J. Amer. Oil Chem. Soc. 39:315. Punwar, J.K. and Derse, P.H. 1978. Application of the official AOAC cholesterol method to a wide variety of food products. J. Amer. Oil Chem. Soc. 61(3):727. Rice, E.E., Poling, C.E., Mone, P.E. and Nayner, N.O. 1960. A nutritive evaluation of over-heated fats. J. Amer. Oil Chem. Soc. 37:607. Rock, S.P. and Roth, H. 1964a. Factors affecting the rate of deterioration in the frying qualities of fats. 1. Exposure to air. J. Amer. Oil Chem. Soc. 41:228. Rock, S.P. and Roth, H. 1964b. Factors affecting the rate of deterioration in the frying qualities of fats. II. Type of heater and method of heating. J. Amer. Oil Chem. Soc. 41:531. 113 Sahasrabudhe, M.R. 1965. Introduction: Studies on heated fats. J. Amer. Oil Chem. Soc. 42:763. Sanche, L. and Van Lier, J.C. 1976. Tracer studies of cholesterol degradation induced by ionized gases. Chem. Phys. Lipids. 16:225. Seher, A. and Homberg, E. 1968. Die untersuchung von sterin-gemischen mit helfe der Dunnschichi-Chromatographic. Fette Seifen Anstrech. 7:481. Sheppard, A.J. and Shen, C.J. 1980. Activities of FDA's division of nutrition regarding cholesterol oxides. In Autoxidation in Food and Biological Systems, M.G. Simic and M. Karel, Eds., Plenun Press, New Y0rk City, N.Y. pp. 133-140. Sims, R.P.A. 1957. Possible mechanisms in thermal polymerization of vegetable oils. II. Polymer formation. J. Amer. Oil Chem. Soc. 34:466. Slover, N.T., Lanza, E. and Thompson, R.H. Jr. 1980. Lipids in fast foods. J. Food Sci. 45(6):1583. Smith, L.L. 1980. The autoxidation of cholesterol in autoxidation. In Food and Biological Systems, M.B. Sim and M. Karel, Eds., Plenum Press, New York City, N.Y. pp. 119-132. Smith, L.L. and Hill, F.L. 1972. Detection of sterol hydroperoxides on thin-layer chromat0plates by means of the Nurster dyes. J. Chromatog. 66:101. Smith, L.L. and Kulig, M.J. 1975. Sterol metabolism. xxx1v. On the derivation of carcinogenic sterols from cholesterol. Cancer Biochem. Biophys. 1:79. Smith, L.L. and Kulig, M.J. 1976. Singlet molecular oxygen from hydrogen peroxide. J. Amer. Chem. Soc. 98:1027. Smith, L.L., Kulig, M.J., Muler, D. and Ansari, G.A.S. 1978. Oxidation of cholesterol by dioxygen species. J. Amer. Chem. Soc. 100:6206. Smith, L.L., Kulig, M.J. and Teng, J.I. 1973a. Sterol metabolism XXVI. Pyrolysis of some sterol allylic alcohols and hydroperoxides steroids. 22:627. Smith, L.L., Matthews, J.C., Price, J.C., Bachman, R.C. and Reynolds, 8. 1967. Thin-layer chromatOgraphic examination of cholesterol autox- idation. J. Chromatog. 27:187. Smith, L.L. and Teng, J.I. 1974. Sterol metabolism. XXIX. 0n the mechanism lipid peroxidation in rat liver. J. Amer. Chem. Soc. 96: 2640. 114 Smith, L.L., Teng, J.I., Kulig, M.J. and Hill, F.L. 1973b. Sterol metabolism. XXIII. Cholesterol oxidation by radiation-induced processes. J. Org. Chem. .38:1763. Smouse, T.H. 1975. The frying properties of a series of vegetable oils and the effectivness of selected tests in predicting their frying stability. Cento panel meeting on regional problems of edible fats and oils. PCSIR Laboritories, Lahore, Pakistan. Strock, H., Ball, C.O., Chang, 5.5. and Steir, E.F. 1966. Effects of agitation and temperature in the deep-fat frying of potatoes. Food Technol. 20:545. Taylor, C.B., Peng, S.K., Nerthessen, N.T., Tham, P. and Lee, K.T. 1979. Spontaneously occuring angiotoxic derivatives of cholesterol. Am. J. Clin. Nutr. 32:40. Teng, J.I., Kulig, M.J., Kan, G. and Van Lier, J.C. 1973. Sterol metabolism. XX. Cholesterol 70 -hydroperoxide. J. Org. Chem. 8:119. Teng, J.I., Kulig, M.J. and Smith, L.L. 1973. Sterol metabolism XXVI. Gas chromatographic differentation among cholesterol hydroperoxides. J. Chromatog. 75:108. Teng, J.I. and Smith, L.L. 1973. Sterol metabolism. XXIV. On the unlikely participation of singlet molecular oxygen in several enzyme oxygenations. J. Am. Chem. Soc. 95:4060. Teng, J.I. and Smith, L.L. 1976. Sterol metabolism. XXXIII. On the oxidation of cholesterol by dioxygenases. Bioorganic Chem. 5:99. Thompson, J.A., Paulose, M.M., Reddy, B.R., Krishnanamurthy, R.G. and Cheng 8.5. 1967. A limited survey of fats and oils commercially used for deep-fat frying. Tsai, L.S., Hudson, C.A., Ijichi, K. and Meehan, J.J. 1979. Quantitation of cholesterol o-oxide in eggs by gas chromatography and big: pggformance liquid chromatography. J. Amer. Oil Chem. Soc. :1 A. Tu, C., Powrie, N.D. and Fennema, O. 1967. Free and esterified cholesterol content of animal muscles and meat products. J. Food Sci. 33:581. Van Lier, J.E. and Smith, L.L. 1967. Sterol metabolism Part l.-26- hydroxy cholesterol in the human aorta. Biochemistry. 6:3269. Van Lier, J.E. and Smith, L.L. 1968. Sterol metabolism. 11. Gas chromatographic recognition of cholesterol metabolitics and artifacts Anal. Biochem 24:419. 115 Van Lier, J.E. and Smith, L.L. 1970a. Autoxidation of cholesterol via hydroperoxide intermediates. J. Org. Chem. 35:2627. Van Lier, J.E. and Smith, L.L. 1970b. Sterol metabolism XI. Thermal decomposition of some cholesterol hydroperoxides. Steroids. 15: Veldstra, H. 1939. 3,5-cholestadiene from cholesterol oleate and its possible bearing on the formation of carcinogenic substances in heated fats. Nature. 144:246. Natt, B.K. and Merrill, A.L. 1963. Composition of Foods. Agr. Handbook No. 8, U.S. Dept. of Agr. Nash. D.C. RICHIGRN STATE UNIV. LIBRRRIES IIIIIIIIIIIIII III IIIIIIIIIIIIII (III (III III 31293104699933