The rise of obesity in America in the last decade of the twentieth century has led to an explosion of research on metabolic consequences of excess adipose tissue. Diet fads that focus on either low-fat foods or low-carbohydrate (and thus high-fat) foods have prompted many studies of the metabolic consequences of dietary fat levels. Remarkably, much of this research has completely ignored the composition of fat and focused only on dietary amount, implying that the composition is unimportant. Many papers, including ones published in highly ranked research journals, report animal studies in which one group is fed a high-fat diet of unspecified composition that induces overeating and obesity and a control group is fed a standard rodent diet that contains a lower amount of fat, again of unspecified and almost certainly different composition.

That so little attention would be given to the composition of the dietary fat is remarkable in light of the large number of human and animal studies showing that dietary fatty acid composition has a profound influence on many aspects of metabolism. Detailed studies in humans and experimental animals, conducted in the 1950s and since, have shown that each fatty acid has unique but overlapping sets of metabolic properties (Warden and Fisler, 2008).

Could changes in dietary fatty acid composition have a role in the obesity epidemic? Do some fats adversely affect human health more than others? Alternatively, does obesity so overwhelm metabolism that the role of fat composition becomes much less significant? Whatever the answer, the issue cannot be overlooked.

The fatty acids consumed by Americans changed dramatically in the twentieth century. Seed oils such as soy, corn, and canola oils were rare before the industrial revolution of the 1800s because they require mechanical crushing or solvent extraction for efficient production. Fruit oils such as olive and palm oils, along with rendered animal fat (lard and tallow), were more widely used. The high-quality taste and low cost of seed oils drove a rise in seed oil production throughout the twentieth century. Today, soybean oil accounts for a staggering 20% of calories consumed by Americans, in the form of mayonnaise, deep frying fat, salad dressings, margarine, nondairy coffee creamers, snack foods, and sandwich spreads. It may be found in any food with an ingredient list that refers generically to “vegetable oil.”

The application of conventional and modern molecular methods (genetically modified) to engineer fatty acid composition of various food oils is resulting in the introduction of oils with modified fatty acid content into the food supply. A notable example is the genetic modification of soybeans to produce oils high in oleic acid. High oleic soy oil was developed to replace the use of trans fatty acid–rich hydrogenated fats for deep frying and other purposes for which an oil with high oxidative stability is needed. The high oleic soybean oil was generated by downregulating expression of the fatty acid desaturase gene that encodes the enzyme that converts monounsaturated oleic acid to the polyunsaturated linoleic acid. The oil from these soybeans contains about 80% oleic acid, compared to 25% for conventional soybean oil. At the same time, it contains less than 9% linoleic acid compared to 54% in conventional soybean oil, and less α-linolenic acid, 3% compared to 7% in conventional soybean oil. Commercial production of these high oleic soybeans was approved in North American countries in 2009–2010. Further modification of the fatty acid composition (e.g., increasing the α-linolenic acid content) is under development. Similarly, high oleic acid peanuts, with only 3% linoleic acid, are already on the consumer market in Australia as peanut butter and peanut-containing snacks. These current and upcoming changes to the fatty acid composition of the food supply will result in a major change, once again, in the fatty acid composition of fats in our diets and, with this change, we will likely see physiological consequences related to fatty acid composition rather than amount of fat.

The odd list of designations we now have for vitamins—A, D, E, and K for the fat-soluble vitamins, several Bs (with a few numbers missing) and C for the water-soluble vitamins, and some with no letters or numbers—is a remnant of the uneven but eventual progress in scientific understanding of the group of essential nutrients we call vitamins. What of other letters between E and K?

Vitamin F was proposed as the name of a factor associated with fat, and this proposal appeared in several papers in the 1920s. The discovery that fat-containing diets are essential for health is usually assigned to a 1929 paper of George and Mildred Burr (Burr and Burr, 1929), then working in the attic of the University of Minnesota medical school (Holman, 1992). They showed that rats fed fat-free diets developed scaly and sore skin, especially on the face and tail; had hair loss; grew at about two thirds of the normal rate; had a shortened life span; and were unable to reproduce. Animals that were emaciated and covered with very scaly skin were placed on 10 “drops” of lard per day. The animals immediately showed signs of recovery and within 10 weeks were fully “cured.” No recovery was seen when the non–fatty acid (nonsaponifiable) fraction of lard was used, thus indicating that the essential fat consisted of fatty acid(s). Subsequent studies established that the fat of greater unsaturation improved the ability of the fat to cure eczema (Brown et al., 1938). However, the inability of scientists to reproduce these effects in humans reduced their interest in vitamin F. We now know that periods greater than 6 months on fat-free diets are required to cause overt deficiency symptoms in humans. Today, the essentiality of the n−6 and n−3 fatty acids, linoleic (18:2n−6) and α-linolenic (18:3n−3) acid, is clearly established, and we would consider these to be components of the so-called vitamin F. There is still controversy about whether humans can synthesize sufficient long-chain PUFAs from 18:2n−6 and 18:3n−3 for optimum health. If, in fact, long-chain fatty acids are essential at some or all stages of the life cycle, PUFAs such as 20:4n−6, 20:5n−3, and 22:6n−3 might also be considered separate components of vitamin F. As progress continues toward defining the details of human requirements for PUFAs, as well as the role of specific PUFAs in human metabolism, the concept of a series of vitamin Fs may yet prove useful.

Trans fats refers to the presence of trans double bonds in an unsaturated fatty acyl group in the fat. Trans double bonds arise from two sources: chemical catalytic hydrogenation of unsaturated oils and as normal products of rumen bacterial and bovine physiology in dairy products. Those arising from hydrogenation of oils have a generally bad reputation as unnatural promoters of heart disease. Food labeling rules implemented by the U.S. Food and Drug Administration in 2006 specified that the content of trans fats be shown on all food labels, and this resulted in many food producers removing trans fats from their prepared foods. In contrast to the trans fats in hydrogenated fats, the trans fats in dairy products are associated with beneficial physiological effects, at least in animal studies.

Chemical hydrogenation adjusts the melting point of oils, turning them into solid fats, by saturating many double bonds with hydrogen and isomerizing others from cis to trans. Saturated and trans fatty acids in oils have higher melting points than corresponding fatty acids with cis double bonds, and thus any oil can have its melting point finely adjusted by adding just the right amount of hydrogen. Hydrogenation lowers the cost of shortening for baked goods because any inexpensive, high-quality, deodorized/flavor-neutral oil available at a particular time can be hydrogenated to make shortening. Furthermore, hydrogenation destroys PUFAs, particularly α-linolenic acid, which is considered to be a major cause of rancidity that limits the life of frying oil and other high temperature–treated foods.

The trans fatty acids resulting from hydrogenation mainly constitute a series of monoenes with trans double bonds distributed at various positions along the hydrocarbon chain, centered at the site of the original cis double bond. Conjugated double bonds of various positions and geometries are also created. These distributions of trans monoene and diene fatty acids are thought to be atherogenic due to their increasing plasma cholesterol in a manner similar to some saturated fats (Hu and Willett, 2002), especially myristic and lauric acids.

In contrast to hydrogenated fats, dairy products have a very specific distribution of trans fatty acids. The most prominent dairy trans fat is a monoene, trans-11-18:1 (vaccenic acid), with smaller amounts of other monoenes with double bonds located at the C4 to C16 positions. Also present at lower concentrations are dienes known as conjugated linoleic acids (CLAs). The most prominent of these is the cis-9, trans-11-18:2 fatty acid, named rumenic acid (Kramer et al., 1998), constituting around 90% of all CLA in dairy products. This fatty acid is a product of the rumen production of vaccenic acid, which is acted upon by a Δ9-desaturase in the cow’s mammary gland (Kay et al., 2004). The conversion of vaccenic acid to rumenic acid also occurs in humans (Turpeinen et al., 2002). The trans-7, cis-9 isomer of 18:2 is made in a similar way but accounts for much less of the total. Rumenic acid has potent anticarcinogenic properties in rats, and a close structural isomer, trans-10, cis-12-18:2, has antiobesity effects in rats (Pariza, 2004). Neither effect has been confirmed in humans, but research is continuing on the possible health benefits of these trans double bond–containing isomers of monoene and diene fatty acids found in dairy products.

Most regulatory definitions of trans fat, such as those of the U.S. Food and Drug Administration and the Codex Alimentarius for international trade, define trans fat as the geometrical isomers of monounsaturated and polyunsaturated fatty acids having nonconjugated carbon–carbon double bonds in the trans configuration. These definitions thus specifically exclude the trans fats (CLA and vaccenic acid) that are found in dairy products and beef.