History of Soymilk and Dairy-like Soymilk Products - Page 3

by William Shurtleff and Akiko Aoyagi

A Chapter from the Unpublished Manuscript, History of Soybeans and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


©Copyright 2004 Soyinfo Center, Lafayette, California

Page 1 | Page 2 | Page 3 | Page 4


The Early Years (1896-1919) . Soymilk had been known in Europe for 30 years before it was first referred to the US in 1896 by H. Trimble in the American Journal of Pharmacology . This was also the first reference to soymilk in English. Summarizing Inouye's (1895) description of tofu making, Trimble wrote, "The liquid filtrate is white and opaque, very closely resembling cow's milk, while the odor and taste remind one of fresh malt." In 1897 Trimble mentioned soymilk specifically in reference to yuba production: "A sort of film forms on the surface of soy-bean milk which in appearance suggests cream." Trimble had never seen or tasted soymilk and he did not refer specifically to its use as a beverage. In 1899 Blasedale mentioned a "milky-white liquid" when describing Prinsen-Geerligs (1895) discussion of tofu making. Between 1909 and 1916 Ruhrah developed a variety of suspended soymilks (he referred to them as "gruels") which he used as the first soy-based infant formulas, and recommended for use when fresh and sanitary milk was not available or when an infant had vomiting or diarrhea. His ingredients were whole (full-fat) soy flour, condensed milk, barley flour, and minerals. In 1916 Sinclair fed a similar suspended and fortified soymilk to 74 infants who had diarrhea; 44 recovered well. These products became the forerunners of America's first commercial soymilk infant formulas and many of the earliest human studies on the nutritive value of soy protein were done on the growth-promoting qualities of such formulas (see Chapter 6).

The first specific mention in the US soymilk patents were granted to Li Yu-ying (1913), Goessel (1913, 1914, 1915), and Monahan and Pope (1915). The latter patent, the first to be granted to US citizens, described the addition of malt extract, cacao, or chocolate to soymilk. In 1918 Melhuish was granted a US patent which was boiled under a vacuum to remove off flavors and treated with milk ripening bacteria. Abbott Labs in Chicago was granted a patent the same year. By 1919 some 42 articles and patents had been published in the US and Europe; the figure reached 66 by 1929, and 116 by 1939 (Smith and Beckel 1946).

The early history of soymilk production in the US is very vague. In 1916 Piper reported that "At this time a factory in New York state as being equipped for the production of soymilk." No details were given. In 1919 Bowers wrote that "Soy milk and soy milk products are being made in the United States now." Again no specifics. In 1923 Kempski noted. "The production of soymilk in America may be of great significance. In New York state in 1916 a large establishment of this type was created. Soymilk doubtless has a great future. The taste can scarcely be distinguished from that of cow's milk." Finally in 1927 Horvath apparently cleared up the mystery; "In the United States some very good breakfast foods and an excellent finely powdered soybean milk powder "Soy Lac" is made by J.A. Chard, Soy Products, 263 W. 12th St., New York City, who has been experimenting for some time with soybeans." It would appear that this company was the first to make soymilk in America. yet it is not known if this was a liquid wet-grind soymilk or simply a finely ground soy flour. And why was the company not mentioned by Piper and Morse in 1923? Nothing else is known of America's first soymilk maker.

1920-1939 . This period saw a strong growth of interest in soymilk and its production. One reason concerned the growing awareness of and problems with diseases transmitted via milk. For example, in the early 1930s, Dr. M.J. Rosenau, Professor of Preventive Medicine and Hygiene at Harvard Medical School wrote, concerning the spread of tuberculosis and typhoid infected milk: "Impure milk conveys more sickness and deaths than perhaps all other foods put together . . . Milk contains a greater variety of infection than any other food" (Kale 1936). Many Seventh-day Adventists voiced the same concerns. Thevenot (1920, 1923, 1925) was granted numerous patents on manufacture of improved soymilk cream. Piper and Morse (123) gave the best information to date in the US or in English on soymilk, including its history, method of preparation, composition, and utilization in East Asia and the West, and preparation of condensed, powdered, and fermented soymilk, plus "vegetable casein" (soy protein isolate). They noted, "Investigations in America and Europe with vegetable milk (soymilk) indicate that it may be successfully used in place of cow's milk in numerous preparations. The milk has been used with good results in bread, cakes, in creaming vegetables, in custards, in chocolate or cocoa, and in milk chocolate. In special therapeutic cases, vegetable milk can also be used successfully in place of animal milk and is said to rank closely to mother's milk in infant feeding."

Important investigations of feeding soymilk infant formulas to babies were conducted by Schloss (1920), Hill and Stuart (1929), and Rittinger and Dembo (1932). Hill and Stuart, of the Harvard Medical School of Pediatrics, developed a fortified suspended soymilk resembling that developed earlier by Ruhrah; they took the concept to Mead Johnson Co., which used it to develop Sobee in 1929, as described below. Interestingly all of these early infant formulas were soy-flour based suspended soymilks. In 1927 what is thought to be the first PhD thesis on soymilk was written by Y.T. Chiu at Cornell. It was reported that soymilk was used for infant feeding and consumed by millions of people in China, but the Cantonese do not drink it.

As soymilk began to be more widely available in the US starting in the 1930s, the number of articles increased (Dittes 1930). Bibliographies on soymilk were published by the USDA and LeClerc in 1934 and 1936, and the dairy industry began to grow concerned about a potential competitor to cow's milk. In 1936, sounding the alarm, Slawson wrote "Baby's Milk from Beans. Glowing News Story of the Soybean that Staggers Imagination," which was published in a dairy magazine. He noted: "The big news for the dairyman . . . lies in the fact that the soybean milk is now being produced in this country on a commercial basis." Soy flour, he added, was threatening to replace milk in bread and pastries and soy casein to replace dairy casein in paints, paper sizings, textiles, and adhesives. Also in the mid 1930s soymilk began to be used as a calf milk replacer (Shoptaw 1936* #1014).

Soymilk production really took root in North America during this period. In 1921 Rouest in France had noted that there was a factory in Canada making calf milk replacer. This may have been Milque Ltd., as reported by Lohse in 1936.* Allergy to cow's milk and lactose intolerance stimulated development of the first soy-based infant formulas in the US. Most of these were made from soy flour. In 1929 Mead-Johnson Co. produced the first commercial soy-based infant formula in America. Called Sobee, it was made from a mixture of full-fat soy and barley flours homogenized with olive oil, had a dark tan color and beany flavor, and contained many complex carbohydrates that led to intestinal gas (flatus) and poor-smelling stools, but in 1929 it was a godsend to infants allergic to cow's milk (Journal of the American Medical Association 1929; Sarett 1976). Note that this soymilk was developed at about the same time that Tso was starting his research on soymilk for infant feeding in China; in 1929 he published a nutritional study of Sobee. Liquid Sobee was introduced in 1954.

America's first filtered (fiber-free) liquid soymilks (unless Soy Lac from 1916 was such) were introduced in the late 1920s independently by two Seventh-day Adventist groups. By late 1929 T.A. Van Gundy in La Sierra, California was producing La Sierra brand canned soy milk (with the beany flavor removed by processing with live steam; see Chapter 36.4), which was widely sold in southern California health food stores. In 1929 Madison College in Tennessee began to make a liquid soymilk for use in their dining halls and sanitarium; it was first sold commercially in 1931 and canned in 1932. It became popular nationwide.

Another early and very popular soymilk infant formula was Mull-Soy. In 1934, while director of allergy research for the Borden Company, Dr. Julius F. Muller developed a soymilk for his own child, who was highly allergic to dairy milk. Muller formed his own company, Muller Laboratories in Baltimore, introduced Mull-Soy commercially in early 1936, and promoted it by his own efforts. It was a liquid suspended soymilk, made by homogenizing together soy flour, water, and added nutrients. In early 1942 Muller's company was purchased by The Borden Company, which continued to promote the product well. A powdered version became available in the late 1940s. In 1934 Dr. J.H. Kellogg was granted the first patent for making acidophilus soymilk; he used this soymilk therapeutically with excellent results (see Chapter 36.2). In about 1933 or 1934 Loma Linda Foods in Southern California began to make a liquid soymilk similar to that developed by T.A. Van Gundy (see Chapter 36.9).

The first cookbook to contain a recipe for homemade soymilk and recipes for its use was Frances Dittes' Food For Life (1935). Other early recipes were given by Van Gundy (1936) and Kloss (1939). All three of these writers were Seventh-day Adventists. Recipes appeared in most soyfoods cookbooks thereafter.

Henry Ford, one of America's great early promoters of soybeans and soyfoods, also did pioneering work with soymilk, of which he was personally very fond. As early as 1934 he served a 14-course soyfoods meal including cocoa with soymilk and soymilk ice cream to the press at the Chicago World's Fair. In the mid 1930s he built a demonstration soymilk plant at Greenfield Village in Michigan near the main Ford Motor Car factory; here he produced several hundred gallons of soymilk a day. After the beginning of World War II, the soy dairy process was commercialized by one of its production workers, Bob Smith. For details on Ford, see Chapter 40 and Ruddiman (1940).

Perhaps America's greatest early soymilk pioneer was Dr. Harry Miller. After living for many years in China, where he and his won Willis started a major soy dairy in Shanghai in 1936 (as described earlier), he returned to the US in 1938. Willis had returned before him and in 1938 began manufacturing canned soymilk in Utica, new York, then later in Washington, D.C. In the fall of 1939 he and his sons began producing Soya Lac brand liquid (fiber-free and filtered) soymilk at a plant that he and his sons built in Mt. Vernon Ohio (see Chapter 36.5). Like Van Gundy, Madison, and Loma Linda Foods, Miller was a Seventh-day Adventist. Soyalac was generally considered to be better tasting and less prone to clog nipples than the various flour-based products, although at least one series of tests showed it to have a significantly lower protein quality as measured by PER (Gyorgy 1962). Miller's plant was purchased by Loma Linda Foods in 1950. Miller also wrote extensively about soymilk, and did much to promote its use throughout the world.

Another early popularizer of soymilk in America was Jethro Kloss, also an Adventist and author of Back to Eden , which was published in 1939. In his book he extolled the health-giving and healing value of soymilk and gave home-scale recipes for two types of soymilk, plus soymilk ice cream, mayonnaise, cheese, buttermilk, cottage cheese, and whipped cream (Chapter 36.6).

1940-1959 . As in Europe, this was a slow period for soymilk, despite the very strong interest in soyfoods in the US during World War II. An exception was Dr. harry Miller, who wrote nine articles and booklets about soymilk between 1940 and 1959, many of them in the Soybean Digest (see Chapter 36.5). In 1957 he wrote a comprehensive review of research on soymilk to date, emphasizing soymilk in infant nutrition, with 102 references. Smith and Beckel's comprehensive "Soy or Vegetable Milk: A Resume and Bibliography" (1946) gave a useful historical perspective, with 124 references to 1944, including 40 patents. Their review covered various methods for improving soymilk flavor including grinding the soybeans with water at temperatures above 90*C, extraction of soy lipids (oils and fats), use of proteolytic enzymes, and extraction with dilute alkali, all techniques which were "rediscovered" during the 1960s and 1970s.

One interesting work written in 1946 was a MS thesis in Home Economics at Cornell University by Edith W. Illick entitled "A Study of Soybean Milk; Its Preparation and Use in Some Chinese and American Recipes." She was the first to use scientific principles in developing a recipe for homemade soymilk. She found that a wet grind and cold milk extraction followed by 15 minutes of simmering at boiling gave the best flavor. A taste panel found that in Chinese recipes, soymilk went best in Almond Cakes and Chicken & Vegetables. American favorites were Chocolate Ice Cream, Cream of Tomato Soup, Spanish Cream, and Baked Custard. Illick noted that although soymilk in China cost only one-fifth to one-tenth as much as cow's milk, it was not used in typical Chinese cooking. She hoped her work would demonstrate how it could be used in place of water in Chinese cooking to add valuable nutrients.

Soymilk feeding experiments during this period were done by Dean (1953*), Glaser and Johnstone (1952*, with Mull-Soy), Collins-Williams (1956*), and Fomon (1959). Fomon did one of the earliest of the "new generation" of studies on soymilk and human nutrition using nitrogen balance techniques. He reported that a soymilk formula (without added methionine) that supplied 6.8% of calories, when fed to infants resulted in growth and nitrogen equivalent to human milk. Many of the nutritional studies that followed are reported in Chapter 6.

During the War the three major brands of soymilk were Mull-Soy, Soyalac, and Madison Soy Milk. Mull-Soy was the most widely used. Sales of all these and other brands increased markedly during the War. A major innovation in soymilk manufacture occurred in 1956, when Worthington Foods introduced Instant Soyamel, the world's first soymilk to be made from soy protein isolate; the latter was manufactured by the Gunther Co. in Illinois. Worthington had introduced a regular (non-isolate) soymilk, Soyamel, in 1954.

1960-1981 . This period, especially after 1965, saw a major expansion of interest in soymilk because of the development of soy protein isolates for use in infant formulas, the increasing recognition of the high quality of soy protein, and the development of non-isolate soymilks with greatly improved flavors.

The first soymilk infant formula based on soy protein isolate was Fortified Instant Soyamel, introduced in 1961 by Worthington Foods. Similar isolate-based infant formulas were introduced in quick succession: Mead Johnson's Prosobee in 1965, Ross Laboratories' Isomil in 1966, Loma Linda Foods; Instagen (later i-Soyalac) in 1966, and Syntex's Neo-Mull-Soy (late 1960s). Most of these new soymilks were fortified with added methionine to make the PER (when measured on rats) 90% of casein, plus added iodine to counteract potentially goiterogenic substances in soybeans, and zinc because of the zinc-binding phytates in soy. The isolate soymilks were whiter and caused less problems with flatulence and loose or malodorous stools than former (especially suspended soy-flour type) soymilks. A list of the world's major soy-based infant formulas is shown in Figure 12.XX. The American Dietetic Association (1980) strongly urged that soymilk used to feed infants or growing children, or in total vegetarian diets, be properly fortified. By the early 1970s soy protein isolates had largely replaced soy flour in infant formulas. By 1973 it was estimated that about 10% of infants in the US were being fed formulas based on soy protein isolate.

Research on improving the quality (flavor) and yield of soymilk was extremely active from 1960-1981, with much of the important US research being done at Cornell University (New York) and at the University of Illinois. In about 1961 the food and technology group at Cornell, under the leadership of Dr. D.B. Hand, and with Hackler, Steinkraus, and Van Buren, became interested in the soybean as a rich source of low-cost protein. Their attention first turned to soymilk. In 1961, at the Peoria conference on Soybean products for Protein in Human Foods Hand et al. presented Pilot Plant Studies on Soy Milk in which they first proposed a method for converting whole, dehulled soybeans, both dry and soaked, into soymilk without removing the okara (insoluble residue). However a fiber-free soymilk they developed got best ratings for flavor and consistency. In conclusion they recommended the "whole-bean" method as having the best protein quality (PER) and yield, and using the least power and labor. In 1966 Wilkins, Mattick, and Hand of Cornell made the breakthrough discovery that the enzyme lipoxygenase (also called lipoxydase) produces the so-called "beany" flavors in soymilk that some people find objectionable, and these off-flavors are not present in the soybean prior to grinding in the presence of moisture. In 1967 the same group showed that by simply grinding the soaked beans together with boiling water and not allowing the temperature of the slurry to drop below 80*C (180*F), the enzyme could be inactivated, yielding a bland and pleasant-tasting soymilk. Small producers around the world are now using Cornell's boiling-water grind method of making soymilk.

Various researchers since 1914 have suggested the importance of removing the crude soy oil if a truly bland soymilk is desired. In 1973 Steinkraus of Cornell University patented such a method using a two-solvent extraction process; the resultant soymilk was given much higher ratings by a taste panel than that made with the boiling water grind. A similar method, used by Asahi Shokuhin in Japan since about 1977, makes soymilk with a very bland, dairylike flavor.

In the early 1970s researchers at the University of Illinois (Nelson, Steinberg, and Wei) began work on developing an improved suspended soymilk from dehulled soybeans. By late 1973 they had discovered a new basic concept, this time using a pre-blanch in water containing a little baking soda rather than the boiling-water grind developed at Cornell. Patents for the process were granted to these three researchers in 1975 and 1977, and assigned to the University of Illinois Foundation. A detailed description was given by the three in 1976. During the following years, many University of Illinois graduate students in the Department of Food Science wrote doctoral theses on the process and helped to improve it. In the mid 1970s Beatrice Foods, licensing the rights to the patent and made 80,000 gallons of soymilk in three flavors, which they test marketed in Illinois. The acceptability was good but the product never got off the ground. Mitsubishi in Japan and Pulse Foods in Taiwan were licensed the patent but were unable for various reasons to establish the product in the marketplace. By 1979 over $250,000 in licensing frees had been earned by the University, and the process was being used by one small soymilk plant each in India and Sri Lanka. The (unlicensable) pre-blanch concept is said to be in wide use in East Asia since the late 1970s. Major advantages of the Illinois process are the high protein recovery and soymilk yield, the good nutritional value (from the fiber), and the reduced flatulence (from the preblanch). Limitations are a somewhat chalky flavor and mouth feel, a definite throat-drying effect from the okara in the soymilk, and (for small manufacturers) the expense of the two large homogenizers that are required, plus the licensing fees. Processes for making excellent ice cream and yogurt using this soymilk have been developed by the three researchers.

Starting in the mid 1970s a number of researchers studied the use of the enzyme alpha-galactosidase to reduce the flatulence factor in soymilk by hydrolyzing the oligosaccharides into simpler sugars (Crocco 1974*; Thananunkul et al. 1976*). In 1976 Drachenberg and Allred at Loma Linda Foods patented a process for making microwave-cooked powdered soymilk; a continuous microwave cooker was developed in 1979 and the process was used to make commercial soymilk reported to have an excellent flavor (Shurtleff 1980).

In 1979 Shurtleff and Aoyagi published Tofu & Soymilk Production , the first book to contain a detailed summary of the research, equipment, and production methods used around the world to make soymilk.

During the 1970s, just as good tasting, low cost, nutritious, and healthful soymilk was gaining popularity worldwide, the image of cow's milk as the "perfect food" began to fade. In the US per capita consumption of cow's milk reached its peak in 1940, then dropped 14% between 1950 and 1978. (Whole milk consumption fell 27% between 1910 and 1978.) In the US soymilk is used primarily in formulas for infants who are allergic to cow's milk protein or intolerant of lactose. Starting in the early 1970s and widely publicized by the United Nations Protein Advisory Group Statement 17 (Feb. 1972) it was realized that some 70% of the world's adult population is affected by lactose intolerance or galactosemia, which causes unpleasant stomach disturbances if animal milks are consumed. Some 50-100% of the population of Africa, Asia, and Latin America and the Middle East, and 75% of American blacks show lactose intolerance, as compared with only 6-19% of North American Caucasians. In some areas, this problem was solved by mixing lactase with cow's milk to transform the lactose into simpler sugars. In addition, some 10% of all babies are allergic to cow's milk; a much smaller percentage are allergic to soymilk (Rackis 1979). While cow's milk contains cholesterol and 3.25-3.5% fat, a majority of which is saturated, and thus implicated as a cause of coronary heart disease. Small amounts of a growing number of chemicals have been found in cow's milk. Penicillin (used for mastitis), tranquilizers, hormones, agrichemicals from fodder and even radioactive strontium 90 from fallout, which is assimilated like calcium and deposited in human bones and mother's milk. All animals (except some humans) stop drinking milk after they are weaned, and it stops being provided for them. Moreover, cows's milk contains more than twice as much protein and almost three times as much calcium as mother's milk, since it is meant to transform a small calf quickly into a 1,700-pound large-boned cow. Moreover, it takes about 4 pounds of grain and soy feed protein to make 1 pounds of milk protein, a fairly inefficient process. Finally, the rapid increase in soft drink consumption after 1960 caused some of the decline in milk consumption, a most unfortunate switch. Nevertheless, milk contains a rich supply of all nutrients except iron and fiber and forms an effective supplement to grain-based diets. In the US dairy products provide 77% of the calcium, 45% of the riboflavin (vitamin B-2), 12% of the protein in the nation's diet (Hetrick 1969). Many Western nutritionists consider milk to be one of the best all-around foods, as does the Indian/Ayurvedic dietary/spiritual tradition; some Yogis use milk as their sole/soul food.

Starting in the mid 1960s there was a major increase in the number of publications on soymilk, particularly on its nutritional value for use in infant formulas and on methods for improving the flavor. It was generally found that properly fortified soymilk infant formulas were equivalent nutritionally to cow's milk. A review of these findings is given in Chapter 6, and in Van Buren et al. (1964), Swaminathan and Parpia (1967), Smith and Circle (1972, 1978), and the proceedings of the various World Soy Protein Symposia of the 1970s.

In 1965 the USDA Food for Peace Program started a pilot program to develop a fortified soy beverage based on soy flour (Soybean Digest, Aug. 1965). A whey-soya drink mix was developed in 1973 by USDA, AID, and private industry in response to a need for skim milk powder in Third World countries. By September 1978 2.7 million gallons of the product had been shipped overseas (Aguilera 1981).

Starting in the 1970s a number of tofu shops began to make fresh soymilk, packaged in Pure-Pak cartons or plastic bottles for local distribution. These included Quong Hop (from 1972), Hinode Tofu Co., New England Soy Dairy (1980), Farm Foods, Swan Foods, Wy Ky, American Foods, Mighty Soy, Hoven Foods, Redwood Valley, Island Spring, Victor Food Products, and White Wave. A good review of this field was Leviton's Making Soymilk in America (1981). These soymilks were mostly sold locally at natural or Oriental food stores in plastic bottles. Their biggest problem was their short shelf life (7-14 days refrigerated). No data on total production were available.

During the 1960s and 1970s thousands of communities grew up around the US and in many other post-industrial societies. A number of these, such as The Farm in Summertown, Tennessee and Blue Mountain Center of Meditation in Petaluma, California, started a soy dairy in the community to produce 20-100 gallons of fresh soymilk daily (at a cost of $0.30-$0.40 per gallon) to be served in the community as a beverage, used in place of dairy milk in cooking, or made into tofu, yogurt, ice cream, or the like. Methods and equipment for producing community soymilk were given in Tofu & Soymilk Production and reprinted in several national periodicals.

An overview of the soymilk industry in the US in 1981 gives the following picture. The market has two major components; the infant formula market and the adult market. Total production was estimated at 33.66 million gallons, of which 95.0% was from infant formulas. The leaders in the infant formula market were Ross Laboratories' Isomel (i) with a 50% market share, Mead Johnson's Prosobee (40%), Loma Linda's Soyalac (7%), and Wyeth Labs' Nursoy (3%). Of the adult soymilks, Loma Linda's Soyagen had 60% of the market (1.0 million gallons) and Worthington's Foods' Soyamel had 40% (Shurtleff 1981).

On 9 February 1981 the US Food and Drug Administration permitted the use of hydrogen peroxide and heat as sterilizing agents for aseptic packaging, opening the door for the Tetra Pak (Brik Pak) technique of beverage packaging. It had been shown that no hydrogen peroxide or suspected carcinogens remained in the package after sterilization. This also opened the door to the new soymilks with new possibilities in America. Vitasoy, introduced primarily into Oriental markets after the legalization of Brik Pak, was being reformulated into a new line of soymilks to suit Caucasian tastes. And The Cola-Cola Company was considering marketing an isolate based soymilk in Tetra Pak, sweetened with fructose or honey and flavored with all-natural coconut, chocolate, coffee, or vanilla, to the natural and health food trades. It may be brand-named Samson.


Page 1 | Page 2 | Page 3 | Page 4