History of Soy Oil Hydrogenation and of Research on the Safety of Hydrogenated Vegetable Oils - Page 2

A Special Report on The History of Soy Oil, Soybean Meal, & Modern Soy Protein Products

A Chapter from the Unpublished Manuscript, History of Soybeans and Soyfoods: 1100 B.C. to the 1980s

by William Shurtleff and Akiko Aoyagi

Copyright 2007 Soyinfo Center, Lafayette, California

Page 1 | Page 2 | Page3

Developments with Hydrogenation from 1940-1980s . The use of hydrogenation increased substantially during World War II as margarine and shortening expanded in usage since whole animal fats were in short supply. The discovery in the late 1950s and early 1960s at the USDA Northern Regional Research Center that linolenic acid was primarily responsible for the undesirable odors generated during frying with soy oil and the subsequent discovery that this linolenic acid and the attendant odors could be largely removed by light selective hydrogenation proved to be a key step in the popularization of soy oil. Lightly hydrogenated, winterized soy oil was introduced into the US in the mid-1960s and contributed to rapid growth of the soy oil market. In 1962 soy oil first captured more than 50% of the US cooking and salad oil market, taking the lead from cottonseed oil (Fig. ??). As noted in Chapter 47, during the 1940s and early 1950s soy oil advanced rapidly to become America's leading food oil, with production of soy oil passing that of cottonseed oil in 1944, that of butter in 1947, and that of lard in 1953. Since hydrogenation played a key role in making soy oil marketable and since soy oil was the most widely hydrogenated US oil, the total use of hydrogenation increased as soy oil utilization increased, which was very rapidly.

During the 1970s there was increasing interest in using antioxidants to replace hydrogenation in soy oil in order to leave more polyunsaturates and essential fatty acids in the oil, reduce losses from high melting point triglycerides during winterization, and eliminate trans fatty acids.

By the late 1970s hydrogenated oils, and especially hydrogenated soy oil, played a major role in the American diet. Kromer (1976) estimated that of the 3.31 million tonnes (7,300 million lb) of soy oil used in the US that year, about 75% (2.48 million tonnes or 5,500 million lb) was partially hydrogenated. In 1974 visible fat consumption in the US was 25.0 kg (54.9 lb) per person. Soy oil supplied 58.7% (32.2 lb) of this total, with partially hydrogenated soy oil contributing 43.9% (24.1 lb) of the total. In 1979 Emken?? and Dutton estimated that in that year about 60% of all edible fats and oils in the US were hydrogenated.

As mentioned, in most of the world's oil processing plants, hydrogenation is the only main oil processing step still done using a batch process. However continuous may well be the wave of the future. Continuous hydrogenation has been practiced commercially by Procter & Gamble for a long time. The basic design of their system was revealed in a US patent issued to Mills et al. in 1950. By the 1970s continuous hydrogenation processes had started to replace batch process in a number of larger facilities. Three basic designs were fixed bed, moving bed, and slurry type?? Hastert (1981) noted that "Depending on the eventual outcome of the current trans isomer and saturated fat controversies, fixed bed hydrogenation could be a technique making it possible to manufacture margarine and shortening having the physiological and nutritional properties determined to be the most healthful." Dutton (1982) foresaw future continuous hydrogenation systems with computer-transducer linked controls and copper catalysts to give better selective hydrogenation.

Whatever the future may hold, hydrogenation has played the key role in making possible and easy the dramatic switch in the American diet from animal fats (lard and butter) to less expensive vegetable oils and fats (shortening and margarine).




Early Concerns (1900-1939) . Almost since the time that hydrogenated oils were introduced in margarine and shortenings in the early 1900s, questions have been raised concerning their safety and nutritional value compared with natural oils and fats. We are including this rather lengthy analysis for various reasons: (1) There is coniderable interest among consumers (especially among the rapidly growing number interested in natural foods and nutrition) and among oil chemists in this subject and in the larger subject of the relative nutritional value and safety of margarine as compared with butter; (2) Most interested consumers are unaware of the vast body of scientific research on this subject; (3) The popular press that has dealt with the subject has generally done so in an unbalanced and unscientific manner; (4) Analyzing a complex problem using the historical approach can demonstrate the power of such an approach; and (5) The understanding gained from from analysis relates directly to the following sections on shortening and margarine.


Shortly after oils began to be hydrogenated for use in foods (margarine and shortening), a vigorous debate began as to the value and safety of these foods. On the whole, the question was resolved in favor of the new products. Almost all of the early criticism of the new products came from German scientists. There were three main areas of concern: (1) the presence of nickel in the finished foods; (2) the digestibility of the hardened fats; and (3) esthetic considerations based on the use of low-grade oils and whale oil, sometimes prepared under unsanitary conditions. There were also other concerns with the safety of these products (to be discussed later), but they were not directly related to the hydrogenation process. Good reviews of these early controversies with complete citations were given by Ellis (1919, 1930) and by Holmes and Deuel (1921).

Early research on the amount of nickel catalyst remaining in hydrogenated fats was done by Lehmann (1909^??), Boemer (1912^??), Normann and Hugel (1913^??), Meyerheim (1913^??), Lehmann (1914), and Sussmann (1915), all in Germany. Tests with a variety of hydrogenated vegetable oils showed that residual nickel levels ranged from 0.07 to 6.3 mg. of nickel per kg of hydrogenated oil. This resulted largely from the formation of nickel soaps from free fatty acids of the oil. It was shown that even at the highest levels of residual nickel and unthinkably high levels of hydrogenated fat consumption, the amount of nickel consumed from the fat would be less than that typically ingested from the consumption of food cooked in nickel-containing utensils, apparently then widely used, and was not likely to exceed 0.6 mg per day. Extensive tests feeding large amounts of hydrogenated fats to various animals and to humans (Lehmann and his family) for more than 6 months caused no observable disorders. These German studies were replicated by Lackey and Sayre (1917) in the US. All concluded that properly prepared hydrogenated oils do not contain sufficient nickel to be harmful to good health. By 1921 the US Federal Meat Inspection Division, which then regulated margarine and shortening production, specified that these hydrogenated products could not contain more than 2 parts per million of nickel, an amount about five times the typical level found in commercial products.

Early research on digestibility by Munk (1889^??) and Arnschink (1890^??) showed that fats with a high melting point were absorbed less effectively than those with a low melting point. Tigerstedt (1907^??) showed that those with a melting point above body temperature were much less digestible than those with lower melting points. Thoms and Mueller (1915) confirmed this finding using various hydrogenated oils. Smith, Miller, and Hawk (1915), in the first US study, found that lard (melting at 45°C) was 94.7% digested by humans while hydrogenated cottonseed oil (melting at 36°C) was 93.4% digested, a difference that was not considered significant. Pekelharing and Schut (1916^??) found that the digestibility of hydrogenated oils was inversely proportional to their melting points and that various animals fed diets in which all of the fat was hydrogenated showed no signs of ill health. Erlandsen et al. (1918^??) found that hydrogenated whale oil and butter had about the same digestibility, more than 91%. During World War I the USDA Office of Home Economics took a strong interest in the digestibility of hydrogenated and unhydrogenated fats. Langworthy and Holmes (1917) found that the digestibility of 52 unhydrogenated animal and vegetable fats ranged from 93-99% They made no reference, however, to soy oil or to hydrogenation. Holmes (1918) found that the digestibility of seed oils ranged from 96.5-98.9%, with unhydrogenated soy oil being 97.5%. In hydrogenation experiments, Holmes and Deuel (1921) and Langworthy (1923), all of the USDA, reported that hydrogenated cottonseed, peanut, and corn oils with melting points of 33-37°C were 94.7-98.1% digestible, but when the melting point was increased by additional hydrogenation to 50°C, the digestibility dropped to about 88.5-92.0%, still quite high. With these studies, the concern over the possible poor digestibility of hydrogenated oils was effectively put to rest. Today it is generally stated that the digestibility of soy oil margarines is typically 96.7% compared with 97-98% for unhydrogenated soy oil.

As early as 1912 Bohm^?? noted that although the "development of oil hardening may mean a great advance technically, it is coupled with such an opportunity for the employment of low-grade raw materials that it is likely to cause anxiety on the part of the public." Hydrogenation does have the capacity to greatly improve the quality of low-grade or partially rancid oils. In a related concern Lewkowitsch (1913), in his authoritative book on oils and fats, reported that the use of hydrogenation-hardened fish liver oils and whale blubber oils in foods was looked upon with disfavor by medical opinion in Germany and Austria. The reasons for this concern are not clear, but apparently these marine oils were not widely used thereafter.

Thus, after considerable investigations of early concerns, a consensus emerged between 1915 and 1925 that hydrogenated oils were not harmful to human health. Moreover, hydrogenation was found to greatly delay rancidification and extend the shelf life of most oils (Knapp 1913^??) and to effectively lower the price of several basic foodstuffs containing hydrogenated oils and fats (Lehmann 1914; Sussmann 1915; Thoms and Mueller 1915). It is perhaps interesting to note that, since soy oil was not yet widely used in foods, none of these early tests on the safety of hydrogenated oils were done on soy oil.

A major area of concern with hydrogenated oils and fats continues to be the trans fatty acid isomers they contain. The discovery and investigation of trans fatty acids traces its origins to the early 1800s. In about 1815 Michel Chevreul in France discovered that natural oils and fats contain many different fatty acids. Then in 1819 J.J.E. Poutet^?? in France isomerized oleic acid to elaidic acid, a semisolid fat resembling pork, by treating olive oil with mercurous nitrate. This is the earliest reported example of conversion from a cis to a trans fatty acid isomer. Poutet called the process "elaidinization," and from it the name of the first known trans fatty acid, elaidic acid, was derived. Yet it was not until many years later that elaidinization was recognized as an example of cis-trans isomerization. In 1832 Boudet^?? also isolated elaidic, and some (Moore 1919) credit him with its discovery. In 1844 Lerch^?? discovered a trans fatty acid in butter. He did not name it, nor did he realize that this was the first trans fatty acid to be found in nature. In 1887 and 1888 Saytzeff^?? first used the term "iso-oleic acid" to refer to what he considered to be a solid isomer of oleic acid; the term continued to be used into the 1950s. In 1919 C.W. Moore, a researcher at Messrs. Joseph Crosfield and Sons, Ltd., in Warrington, England (the first company to attempt commercial hydrogenation), made an extensive examination of the whole question of the formation of solid iso-oleic, elaidic, and trans acids during hydrogenation. He first described the migration of double bonds in oleic acid isomers and showed clearly that the formation of isomers during hydrogenation is a major cause of the hardening phenomenon. In 1928 Bertram in the Netherlands found small amounts (0.01%) of a trans fatty acid (11 t- 18:1), which he called vaccenic acid (after vacca = cow), in the fats of ruminants, specifically those in cattle, sheep, and butterfat. He noted that vaccenic acid was a close relative of elaidic acid and that it was the first trans fatty acid to be found in nature, and in foods widely consumed by humans. In 1929 Hilditch and Vidyarthi in London showed that partial hydrogenation of oleic acid and its esters resulted in the formation of numerous positional and geometric isomers of oleic acid; the natural cis configuration of the double bonds had in many cases been converted into a trans configuration. Also in 1929 Hilditch and Jones^?? reported the presence of isomers of linoleic acid in butterfat. It was then postulated that these were a mixture of cis-trans and trans-cis -9, 12 isomers of linoleic acid. Also in 1929 Burr and Burr discovered that some fatty acids were essential for good health; the main essential fatty acid was later found to be linoleic acid. In 1930 Grossfeld and Simmer^?? in Germany reported 1.49% vaccenic acid in margarine. In 1930 Ellis, in his book Hydrogenation of Organic Substances , reported commercial shortenings contained 30% or more of an isomer or isomers of oleic acid (then called "iso-oleic acid" but now called elaidic acid) which was then thought to consist largely of elaidic acid (?? 9 t- 18:1). Ellis discussed the composition of iso-oleic acid. In 1933 Barbour in Ontario did the first research on the metabolism of trans (iso-oleic) fatty acid, which, he noted, was not known to occur as a constituent of any natural fat. He fed a diet containing 40% hydrogenated cottonseed oil (itself containing 26% iso-oleic acid) to rats and concluded that "iso-oleic acid may take part in the fat metabolism of the body without any greater difficulty than any other dietary fatty acid, and that it is in no way objectionable as a constituent of foodstuffs." Barbour reported that, like any other fat, this trans fatty acid was absorbed by the body, deposited unchanged in the body fat, then metabolized as readily as any other fat. It disappeared during fasting. Trans fatty acids had apparently passed the first investigation into their safeness in animal diets. Likewise Sinclair (1935) used elaidic acid (not produced by hydrogenation) as a labeled compound and showed that it was deposited in rats, replacing 25-30% of the saturated fatty acids of liver and muscle phospholipids. This seemed to indicate that it was metabolized like a saturated fatty acid. it entered and disappeared from the liver rapidly, from muscle slowly. Thus at the end of the 1930s there was no good reason to suspect the safeness and nutritional value of hydrogenated fats.

1940-1959 . Little research was done on fatty acid isomers during the 1940s, since World War II offered more pressing subjects for study. Starting in 1941, with the outbreak of war in the Pacific Ocean, the basic US margarine formula was changed, with highly saturated, imported coconut oil being replaced by hydrogenated domestic soy and cottonseed oils. This change resulted in a marked decrease in the saturated fatty acid content (Alfin-Slater et al. 1957), and also led to the first carefully controlled, long-term feeding tests of hydrogenated fats in order to carefully re-examine the safety of the former. In 1945 Deuel and co-workers at the University of Southern California reported on the first multi-generation rat feeding study in which hydrogenated vegetable fat (margarine) supplied virtually all of the fat in the diet. Ten generations of rats continued in excellent nutritional condition, with the growth rate of each successive generation increasing, and excellent fertility. In 1950 Deuel (now one of the world's foremost authorities on oils and fats) and co-workers reported that the 11th through 25th generations were equally healthy in terms of growth, maturation, and reproduction. By the 1940s a new development, selective hydrogenation started to be practiced widely. Bailey (1949) noted that this process led to greater production of fatty acid isomers.

The 1950s were a time of extensive research on fat metabolism and fatty acid isomers for at least three reasons: (1) New techniques, especially spectroscopy and gas chromatography, provided researchers with new tools for rapidly and accurately analyzing complex mixtures of fatty acids and lipid-related compounds. Scientists learned that fatty acids occur in nature in far greater variety than previously expected and that many have unique biological properties; (2) The so-called Lipid Hypothesis, stating that the level of serum lipids (especially cholesterol and triglycerides) was an important risk factor in coronary heart disease, seemed to provide hope that this major disease might be understood and conquered. The consumption of polyunsaturated fatty acids (PUFA) was shown to lower serum lipids, yet hydrogenation was known to lower the percentage of PUFA and essential fatty acids in an oil. Could, therefore, hydrogenation and trans fatty acids be a factor in promoting coronary disease? (3) The consumption of hydrogenated fats in the form of margarine, shortening, and lightly hydrogenated oils (especially soy oil) increased rapidly during the postwar period; (4) Selective hydrogenation often increased trans acid levels.

Various studies during 1950 and 1951 showed that commercial products contained a minimum of 4% and a maximum of 40% trans iso-oleic acid, depending on the conditions of hydrogenation. Jackson and Callen (1951^??) found at least six different trans iso-oleic acids in partially hydrogenated soy and cottonseed oils. New interest developed in the trans fatty acids found in animal fats. In 1951 Reiser^?? showed that bacteria in the rumen of ruminates (cattle, sheep, etc.) hydrogenate linolenic acid to linoleic, which is presumably the source of the traces of trans acids found in the fat of ruminants, as reported by Bertram in 1928. Then in 1952 Swern and co-workers found that beef fat contained not traces but 5-10% of trans -acids. These consisted of a mixture of elaidic and vaccenic acids, both trans isomers of oleic acid (18:1). A year later, by a similar procedure, Cornwell and co-workers found 5.0-9.7% trans -acids in summer buttermilk fat. In 1954 Hartman and Shorland summarized this work, noting that trans -acids were not found in the fat of non-ruminants fed on plants (0.9% was found in the fat of a pig fed skim milk) but were found in the fat of all plant-fed ruminants tested (ox, sheep, deer) at levels of 3.5-11%. In 1957 Johnston, Johnson, and Kummerow at the University of Illinois published the first report on the occurrence of trans fatty acids (2.3-12.2%) in human tissue, where they were deposited presumably after consumption of foods (ruminant fats, hydrogenated fats) containing trans . In 1959^?? Kuehn and Luck also reported 3-4% trans fatty acids in human fat. And in 1977^?? Picciano and Perkins reported 2-4% trans 18:1 acids in the fat of the milk of human mothers.

The discovery of essential fatty acids by Burr and Burr in 1929 and their coining of that term in 1930 led to great interest in this subject during the 1950s. In 1951 Holman and in 1955^?? Privett and co-workers showed that although linoleic acid is the main essential fatty acid, the trans isomers of essential fatty acids produced by hydrogenation are converted into inactive forms which no longer function as essential fatty acids. It also seems probable that these isomers are metabolized through different pathways. In 1954 Deuel^?? observed that hydrogenated coconut oil and fats rich in elaidin accentuated fat deficiency symptoms. While such symptoms were rarely a dietary problem, the concept became important in the future in explaining problems which arose in test diets containing hydrogenated fats but deficient in essential fatty acids (EFA). In 1954 Melnick and Deuel, in a comprehensive report, found that while hydrogenation lowered markedly the EFA content of vegetable oils, the resulting hydrogenated products were still somewhat better sources of essential fatty acids than natural butter oils. Holman and Aaes-Jorgensen (1956) found that trans isomers aggravated the problems in an EFA-deficient diet. In 1956 Sinclair at Oxford, in a long letter to The Lancet , argued that "during hydrogenation much of the EFA are destroyed and unnatural trans fatty acids are formed," and that this was contributing to a chronic deficiency of EFA, which he felt was a major cause of coronary thrombosis and cancer, based largely on epidemiological evidence. In the mid-1950s linoleic, linolenic, and arachidonic acid were all considered essential fatty acids. In 1957 Alfin-Slater et al. put the subject of EFA deficiency into perspective by noting that butterfat contains less than half of the EFA concentration found in hydrogenated margarine oils.

During the next 2 decades it was recognized that EFA deficiencies in human diets are rare since the EFA requirement was considered to be about 2% of the daily calorie intake, or about 4-5 gm per day. Aaes-Jorgensen (1961^??) was among the first to note that contentions that high levels of trans fatty acids increase the need for EFA were often if not always based on experiments with animals deficient in EFA, a key point. In 1977^?? FAO revised its EFA requirements upward to 3% of total calorie intake for normal adults and infants, to 4.5% during pregnancy, and to 5-7% during lactation. An EFA intake of 10-13% of calories was recommended as a precaution against cardiovascular disease. However subjects have been fed less than 2 gm of linolenic acid per day for 6 months without showing any adverse effects. In the US fats typically comprise 42% of total calorie intake and an estimated ?? % of these (??% of the total) are EFA, far more than the minimum recommended amount. Recall, however, that trans -containing polyunsaturated fatty acids do not function like essential linoleic acid and may actually increase slightly the requirement for essential fatty acids (Jensen 1976^??; Privett et al. 1977^??; Erickson?? 1980, p. 448).

Another important line of investigation concerned the relationship between hydrogenation, unsaturated fatty acids, trans fatty acids, and coronary heart disease. Various reports in the early 1950s (Kinsell et al. 1951^??; Ahrens et al. 1955; Kinsell and Michaels 1955^??; Bronte-Stewart et al. 1956^??) stated that high blood cholesterol levels could be reduced by ingestion of fats containing a high proportion of unsaturated or polyunsaturated fatty acids. Bronte-Stewart et al. reported that the cholesterol-depressing effect of highly unsaturated oils was lost after hydrogenation. Ahrens et al. (1957) noted generally higher blood cholesterol levels in human subjects when unsaturated oils were replaced in the diet by their hydrogenated counterparts, although results were not uniform. They were emphatic in pointing out that lowering of serum lipids led to decreased incidence of coronary heart disease.

Sreenivasan and Brown (1956) found commercial margarines and shortenings to contain 25-40% trans isomers of oleic acid, 2-8% linoleic acid and "considerable proportions of both 9, 12- cis,trans or trans,cis and isolated cis,trans isomers of linoleic acid." Mabrouk and Brown (1956) in an analysis of five commercial shortenings and six margarines, found that they contained an average of about 30% (range 22.7-41.7%) trans fatty acids, and 0.2-1.5% conjugated diene. Noting that Americans were then consuming nearly 1,000 million lb or 453,000 tonnes (5.92 lb or 2.69 kg per capita) of trans fatty acids each year, they concluded: "It is indeed fortunate that at present there is no evidence to indicate that these unique acids are in any sense deleterious. The principal disadvantage of hydrogenation is that it removes large amounts of essential fatty acids which are present in the natural oil."

An update on the now-classic multi-generation rat feeding study reported on earlier by Deuel et al. (1945, 1950) was presented in 1957 by Roslyn Alfin-Slater and co-workers from the Department of Biochemistry and Nutrition at USC. Deuel had died in 1956. The rats had now thrived through 46 generations on a ration in which virtually all of the fat was supplied by a selectively hydrogenated margarine composed of equal parts soy and cottonseed oil. The fat level in the diet increased over the years from 9.2-11.2%, which represented about 21-25% of the calorie intake. The margarine contained 35.3% trans fatty acids, 19% saturated fatty acids, and 3.7% essential fatty acids. (After generation 44 the EFA content was increased to 6.9%). Calculations showed that the rats ingested about 10 times as much trans fatty isomers per unit of body weight as would be ingested by humans on a diet having all of the visible fat in the form of selectively hydrogenated vegetable oils. The rats continued to do well, as measured by gain in weight, tibia length, reproduction and lactation, longevity, and histopathological examination of the tissues. The authors concluded that "selectively hydrogenated vegetable oils, such as are employed in margarine manufacture, containing positional and stereoisomers of the unsaturated fatty acids, are fully digestible, harmless, and full of nutritional value as determined by long-term studies conducted with rats."

In 1973 Alfin-Slater et al. reviewed the work of 32 years of multi-generation studies on rats fed margarine of varying amounts and compositions and again found no problems. One might have expected further research and debate on the safety of fatty acid isomers to have subsided in 1957 (Applewhite 1981), but actually the issue was just starting to attract attention. Emken (1959 ref??) noted one possible weakness in the experiments' design: "the life span of common laboratory animals may be too short to allow slow cumulative effects to develop."

One study in 1958 by Johnston, Johnson, and Kummerow basically confirmed the findings of Barbour in 1933. The 1958 study concluded that rats were capable of efficiently metabolizing trans fatty acids over a 3-month feeding period. When the trans acids were removed from the diet, they gradually decreased in amount in the tissues. The presence of trans acids in the diet did not appear to inhibit growth. In fact at the end of the 1950s, as was the case 20 years earlier, there seemed to be little scientific evidence that hydrogenated oils and fats were anything but healthful and nutritional.

1960-1980s . Starting in the early 1950s, popular nonscientific writers began for the first time to revive old debates about the safety and nutritional value of hydrogenated oils and fats. Once again the concern over the lack of digestibility of hydrogenated fats was hotly discussed, but this time not in a scientific context (Bailey 1951). During the 1960s, as concern with heart disease increased, the popular press gave the matter greater attention. One of the most strident and emotional treatments of the subject was Margarine: The Plastic Fat and Your Heart Attack , published in 1962 by J.H. Tobe of Provoker Press. Tobe had read and he quoted several articles showing that hydrogenation in oils can raise serum cholesterol levels and reduce linoleic acid content of oils, but he made no attempt to give a balanced picture and discuss both sides of the issue. He felt that hydrogenated fats were . . . "unqualifiedly the greatest known threat to the health of the people at the present time . . . Hydrogenated fat is nothing but sheer, hard plastic. It will not melt when you squeeze it or rub it against your fingers." Tobe was concerned about the less than 0.6 parts per million of nickel that he said remained in hydrogenated fats, and about the fact that there were no laws requiring the listing of hydrogenated fats on food labels or limiting the degree to which food fats can be hardened. If Tobe and other nonscientific writers had taken the time to do their research carefully, they would have discovered that the concerns which they were raising had already been raised and investigated from 1907 to 1920, and had been shown to be without substance. Tobe's one-sided, unscientific and sensationalist anti-hydrogenation attitude was to characterize most of the popular health-food and natural food literature (mainly magazine articles) on the safety and health value of hydrogenated oils into the 1980s. The problem was not only that his conclusion contradicted the weight of scientific evidence available at the time, but that he made no attempt to be objective. On the other hand, most of the general popular press liked margarine for its low cost and richness in polyunsaturated fatty acids, and, except for a few instances in the 1970s, showed little concern about its safety.

The question of the relationship between hydrogenated oils, serum cholesterol levels, and coronary heart disease, introduced in the early 1950s, was studied in more depth after 1960. One irony of the issue was that just as researchers were debating whether or not hydrogenation of an oil caused it to raise serum cholesterol levels, margarine makers were starting to promote margarine for its high P/S ratio which was known to lower serum cholesterol. The apparent contradiction becomes clear when we recall that even if hydrogenation does cause an oil to be slightly more saturated than its unsaturated counterpart, margarine still has a much higher P/S ratio than butter.

In 1960 Horlick^?? was the first to suggest that trans fatty acids formed during hydrogenation probably raised blood cholesterol levels. The first supportive evidence of this idea was given by Anderson and co-workers in 1961. They showed that men consuming trans fatty acids had significantly higher serum cholesterol levels and concentrations of triglycerides than men consuming the same type and amount of fat but without trans fatty acids. Butterfat, however, produced the highest levels of serum cholesterol and phospholipids. Similar results were reported that same year by Keys, Grande, and Anderson^??. In 1962, however, McOsker et al. in an elegant human feeding study, reported contradictory results, namely that partially hydrogenated vegetable oils had no significant effect on serum cholesterol levels. However, of the diets that contained partially hydrogenated fat, the one with the highest percentage of trans fatty acids also had the highest serum cholesterol level. A similar study by Erickson (1964^??) reached the sa e conclusion as McOsker, but the percentage of trans fatty acids in the partially hydrogenated soy oil was small and the percentage of linoleic acid was high. Another well controlled study by DeIongh?? et al. (1965^??) demonstrated the striking effect of essential fatty acids on serum cholesterol and could not substantiate the suggestion of Anderson et al. (1961) that trans had an effect comparable to saturates. Yet Rand and Quackenbush (1965) showed that rats fed purified trans fatty acids had higher serum cholesterol levels than those fed their cis counterparts. Vergroesen and Gottenbos (1973^??) showed a cholesterol-raising effect of trans fatty acids fed with an egg. Mattson et al. (1975; from Procter & Gamble) found that diets high in saturated fats with high trans levels did not raise plasma cholesterol levels in 17 adults compared with a control group consuming equal amounts of unhydrogenated fats. Heckers et al. (1977) obtained similar results but suggested that the total trans fatty acid content may not be as important as the quantity of the individual isomers present. In all of the above experiments the raising or non-raising of serum cholesterol depended largely on the balance in the diet of factors with opposing effects??; saturated fats, cholesterol, total fats and calories, total and types of trans fatty acids, and polyunsaturated fatty acids. The significance of the results depended largely on the degree to which serum cholesterol levels were changed.


Page 1 | Page 2 | Page3