Journal of Nutrition Animal Diets/Enrichment Products...

Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
 QUICK SEARCH:   [advanced]


     


This Article
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Carpenter, K. J.
Right arrow Articles by Olson, R. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Carpenter, K. J.
Right arrow Articles by Olson, R. E.

The Journal of Nutrition Vol. 127 No. 5 May 1997, pp. 1017S-1053S
Copyright ©1997 by the American Society for Nutritional Sciences

Experiments That Changed Nutritional Thinking1

Kenneth J. Carpenter2, Alfred E. Harper*, and Robert E. Olsondagger

Department of Nutritional Sciences, University of California, Berkeley, CA; * University of Wisconsin, Madison, WI; and dagger  University of South Florida, Tampa, FL

FOOTNOTES

The objective of this two-part symposium, begun in 1995 and continued in 1996, was to describe some of the discoveries made during the past 150 years that changed the direction of thinking in nutrition. These discoveries all illustrate the strength of the scientific method as a process for gaining reliable knowledge of the natural world.

Philosopher of science Karl Popper proposed that the scientific method begins, not with the accumulation of facts, but with recognition of an unsolved problem. This leads to conjecture about a solution, i.e., formulation of a hypothesis. The essence of the process is to subject hypotheses to critical examination and experimental tests that have the potential to refute them. It is basically a process for detecting error; its strength lies in its self-correcting nature. If a hypothesis fails to withstand a test with the potential to refute it, it must be discarded or modified. It is equally important, nonetheless, to defend hypotheses vigorously to ensure that they are not rejected without being tested thoroughly. Although the ability of a hypothesis to withstand such tests does not establish unequivocally that it is valid, as assumptions that are false are eliminated by repeated testing, we achieve an increasingly better approximation of reality.

The process is illustrated by two articles on early studies of protein that were included in the symposium. During the 1820s, protein was accepted as an essential nutrient on the basis of feeding studies with dogs. Subsequently the German school of Liebig and Voit postulated on theoretical grounds that protein was the source of energy for muscular work. This hypothesis was challenged by Fick and Wislicenus in 1866 in an elegant nitrogen balance study performed during a mountain climbing expedition. Their results, interpreted by Frankland, proved conclusively that the hypothesis was erroneous (Paper 2). Nonetheless, the assumption that a high protein intake was uniquely important in stimulating vigor of mind and body was accepted for another 40 years until it was challenged in 1904 by Chittenden, who demonstrated that healthy young men remained vigorous with a protein intake about half that recommended by Voit (Paper 5).

Thomas Kuhn, another philosopher of science, has concluded that major advances in science do not occur gradually, but suddenly, and constitute "scientific revolutions." He uses the term "paradigm" to describe the theoretical assumptions, laws and techniques that dominate scientific experimentation by a particular community of scientists during a given period. Eventually, however, observations that are at variance with the current paradigm are encountered. The paradigm is recognized as being inadequate, and a new and radically different hypothesis is proposed as the result of unusual insight, usually coupled with new methods. This leads to a new paradigm, and a period during which it is consolidated follows. An example given by Kuhn of a scientific revolution is the discovery by Copernicus, at a time when essentially all astronomers believed that the earth was the center of our solar system, that in fact the sun was the center of the solar system and the planets, including Earth, revolved around it. The history of the various sciences, Kuhn proposes, is characterized by a series of paradigms interspersed with periods of "normal science" during which problems falling within the limits of the prevailing paradigm are explored. This view is complementary to that of Popper, who emphasized the need for constant hypothesis testing and modification to ferret out error. Scientific revolutions have occurred in biology and medicine, of which nutrition is a part, since the time of Hippocrates.

Some examples of scientific revolutions in biology and medicine are the discoveries of Harvey, Lavoisier and Darwin, each of which made existing paradigms obsolete. Harvey, in 1628, discovered that blood pumped by the heart through the arteries passed to the veins and circulated back to the heart. This was the demise of the hypothesis, postulated by Galen in the 2nd century, that the blood oscillated back and forth within the arterial system. Lavoisier's discovery, in 1777, that combustion was a chemical process in which oxygen combined with other elements with the release of energy made untenable the century-old hypothesis that combustion represented loss of "phlogiston." Charles Darwin in his classic study The Origin of Species, published in 1859, assembled evidence that new species had evolved continuously over millions of years. His theory of evolution demonstrated that biblical creationism, the belief that species arose intact through supernatural intervention, and which was almost universally accepted in countries that had adopted Western religions, was incompatible with scientific observations.

All of the symposium presentations that follow discuss experiments that influenced nutritional thinking. Some describe experiments that challenged accepted concepts and resulted in their displacement with new ones; others are reports of discoveries that arose from exploration of specific aspects of the new concepts. Several fit Kuhn's concept of scientific revolutions that bring about a rapid change in the paradigm of a field.

A major paradigm shift in nutrition was the discovery of the essentiality of organic and inorganic micronutrients. Despite a number of observations during the 19th century that diets composed of purified food constituents did not support growth or even life, this shift did not occur suddenly as the result of a single discovery; it occurred over a period of more than 60 years. The lag was attributable in large measure to resistance to the new paradigm by many scientists who were influenced by the great prestige of Liebig and who accepted, almost as dogma, his concept that energy sources, protein and a few minerals were the sole principles of a nutritionally adequate diet. Only after the inadequacy of Liebig's hypothesis had been demonstrated in many experiments that should have changed nutritional thinking, but did not, was the new paradigm generally accepted. Four of the papers describe experiments that contributed to the shift in paradigm.

Gerrit Grijns, in the 1890s, extended the work of Eijkmann in Java (Indonesia) showing that chickens fed a diet of white rice developed polyneuritis, a disease resembling beriberi. The disease was prevented by including rice polishings or beans or water extracts of them in the diet. He concluded that chickens needed an organic complex provided in adequate quantities by rice polishings and beans but not by polished rice. His observations had little immediate effect on orthodox nutritional views, even though they ultimately contributed to the basis for the new paradigm (Paper 4).

Liebig's concept that the nutritional value of foods and feeds could be predicted from their proximate composition (nitrogen, ether extract, ash, and carbohydrate by difference) was tested directly by Hart and colleagues in 1907. They found that calves from cows fed an all-wheat ration survived only a short time even though the wheat ration was balanced for major nutrients to match an all-corn ration that proved to be fully adequate. This was a clear demonstration of the inadequacy of Liebig's concept (Paper 6).

Subsequently, McCollum found that rats fed a simplified diet of casein, carbohydrate and minerals stopped growing unless supplied with a fat-soluble factor present in butter but not in olive oil. Rats fed a polished rice diet were found to need a water-soluble factor B, as Grijns had shown, as well as the fat-soluble factor A (Paper 7). During this period, Holst and Froelich in Norway induced a scurvy-like disease in guinea pigs by feeding them diets resembling those of Grijns. This disease was prevented by providing the guinea pigs with lemon juice or cabbage.

Also, between 1909 and 1914, Osborne and Mendel at Yale, following on an earlier observation by Hopkins in Cambridge that tryptophan was essential for the survival of mice, discovered that some plant proteins did not support growth of rats unless the rats were supplemented with other amino acids (Paper 8). Hopkins, and Funk in London, both postulated in 1912 that diseases such as scurvy, beriberi and rickets were dietary deficiency diseases. Only between 1910 and 1915, after these and other demonstrations of the inadequacy of Liebig's concept, was the new paradigm of the essentiality of minor constituents of foods widely accepted.

Acceptance of the new paradigm was followed by a period of unparalleled discovery in nutritional science from about 1915 to the 1950s, during which some 40 essential nutrients were identified and characterized and their functions explored. Several of the articles included in the symposium discuss representative experiments of this expansion of knowledge within the new paradigm.

Iron was known early in the 19th century to be a component of hemoglobin, but the belief that only organically bound iron was available to the body was an obstacle to understanding the role of minerals in nutrition. The demonstration by Stockman in 1893 that inorganic iron was used efficiently for hemoglobin synthesis corrected this erroneous assumption (Paper 3). Thirty-five years later, Hart and associates discovered that copper was essential for the utilization of iron in hemoglobin formation. It is now known that copper promotes uptake of iron by transferrin and increases the utilization of iron by erythroblasts for hemopoiesis (Paper 10).

After the discovery that yellow carotenoid pigments and colorless oils both had vitamin A activity, a conflict between competing hypotheses about the nature of vitamin A precursors was resolved by Thomas Moore, who in 1930 showed that the yellow beta -carotene was converted to colorless vitamin A in the animal body (Paper 11). Thiamin was shown by Lohmann and Schuster in 1937 to be a component of the coenzyme thiaminpyrophosphate, and its role in pyruvate metabolism in the animal body was elaborated by Peters (Paper 12). Observations by Goldberger that protein as well as protein-free extracts of yeast could cure pellagra posed a problem that was resolved when Krehl and colleagues discovered in 1945 that the amino acid tryptophan was a precursor of niacin in the body (Paper 15). That complex interactions and antagonisms can occur among trace minerals was discovered by Dick and associates, who observed that copper deficiency occurs in animals with a normally adequate intake of copper if their intake of molybdenum and/or sulfate is high (Paper 16). In 1972, selenium was shown by Rotruck and co-workers to be essential for the action of glutathione peroxidase (Paper 20).

Also during the 1970s, through the work of Kodicek in Cambridge and Deluca in Wisconsin, the prevailing view that vitamin D acted directly to promote intestinal absorption of calcium and regulation of bone metabolism was shown to be in error. They discovered that vitamin D, through the combined actions of the liver and kidney, was converted to a hormone that mediated the actions attributed to vitamin D. This represented a new concept: the action of a vitamin depending on its conversion to a hormone (Paper 19).

Another major paradigm shift in nutrition resulted from discoveries about the ability of the body to synthesize and degrade nutrients and tissue constituents. The shift occurred in phases, two of which are discussed in the symposium.

Claude Bernard, the great French physiologist, conjectured about the source of glucose in the blood of dogs consuming a diet that contained neither sugar nor starch. By a series of carefully conducted experiments during the 1850s, he discovered liver glycogen and the process of gluconeogenesis by which glucose and glycogen could be synthesized in the liver from non-glucose precursors, enabling this organ to supply glucose to the blood (Paper 1).

The use of isotopically labeled compounds by Schoenheimer in the 1930s to follow the metabolic fate of fatty acids and amino acids administered orally revealed for the first time that these nutrients were incorporated rapidly into depot fat and body proteins, respectively, and that their metabolites continued to be excreted over many days. Through his work, the concept of distinct exogenous (dietary) and endogenous (tissue) metabolism was replaced with the concept of the "dynamic state of metabolism," the continuous breakdown of tissues with the constituents of both food and tissues entering a common pool from which new tissue components were synthesized (Paper 14).

The demonstration by Becker and colleagues that sucrose and fructose are toxic to young pigs and calves represents an extension of this paradigm, one of many, illustrating that metabolic pathways for some nutrients may not be functional at birth and undergo development during the early stages of growth (Paper 18).

With the successive discoveries of essential nutrients between 1915 and 1950 and the virtual disappearance of dietary deficiency diseases, emphasis in nutrition was on ensuring that diets would provide adequate quantities of all essential nutrients to prevent impairment of growth and development. Although it was recognized that requirements declined with increasing age, little attention was given to the long-term effects of total food intake. One of the first challenges to the paradigm that if essential nutrient intake was adequate throughout life, other dietary factors would be of little consequence, came from Clive McCay. He argued that short-term trials with the emphasis on rapid growth did not provide an adequate test of the most desirable nutritional state throughout life. He found that, although rats allowed to freely eat a nutritionally adequate diet grew most rapidly, those allowed only restricted amounts of food could survive much longer (Paper 13). Competing hypotheses about the basis for these effects remain unresolved, but they have opened new directions in nutritional thinking, especially in relation to appropriate body weight and energy intake for adults.

Emphasis on the paradigm of nutritional essentiality also distracted attention from investigations of the nonnutritional components of foods and from the ancient paradigm that foods contain nutriment, medicines and poisons. The finding that broccoli in the diet increased resistance of guinea pigs to X-irradiation and that this effect was not related to its contribution of known nutrients shifted attention back to the nonnutrient components of foods (Paper 17). There is now evidence that substances in cruciferous plants and some other foods may increase resistance to cancers. These observations have led to acceptance of a scientifically based form of the paradigm that foods can affect health by their contributions of chemicals other than essential nutrients through their influence on susceptibility to certain diseases.

The work reviewed here illustrates how much has been learned through the use of animal models. It also illustrates that caution must be exercised in extrapolating findings in one species to another. For example, rats were an excellent choice for studies of vitamin A and thiamin deficiencies, but failure to produce the equivalent of either pellagra or scurvy in this species led a leader in the field to conclude that these diseases in humans were not, after all, due to dietary deficiencies. These experiments also illustrate the need for caution in assuming that observations made at one stage of life apply throughout life.

The history of nutrition illustrates that new paradigms and concepts do not necessarily make earlier ones obsolete; several may exist together and overlap, with all being valid frameworks for investigation. What is the outlook for new paradigms and concepts in nutritional science? Application of techniques from genetics and molecular biology to nutritional problems has led in recent years to advances in understanding the roles of nutrients and their metabolites in the regulation of gene expression with respect to metabolic adaptations, the action of hormones, and responses of the immune system. Undoubtedly other new paradigms and concepts, unanticipated now, will follow.

We believe that these proceedings illustrate, on the one hand, the tremendous advances resulting from the scientific approach to nutrition and, on the other, the importance of continually maintaining a critical approach to even well-accepted hypotheses and concepts.


FOOTNOTES

1   Presented as part of the minisymposium "Experiments That Changed Nutritional Thinking" given in first part at Experimental Biology 95 on April 11, 1995 in Atlanta, GA, and in second part at Experimental Biology 96 on April 16, 1996, in Washington, DC. This symposium was sponsored by the American Society for Nutritional Sciences. Guest editors for the symposium publication were Kenneth J. Carpenter, University of California, Berkeley, CA, Alfred E. Harper, University of Wisconsin, Madison, WI and Robert E. Olson, University of South Florida, Tampa, FL.
2   To whom correspondence should be addressed.


Paper 1: The Liver Forms, Stores and Secretes Glucose (Claude Bernard, 1860)

Presented by Patricia B. Swan, Department of Food Science and Human Nutrition, Iowa State University, Ames, IA 50011 as part of the minisymposium "Experiments That Changed Nutritional Thinking" given at Experimental Biology 95, April 11, 1995, in Atlanta, GA.

In 1834, the 21-year-old Claude Bernard left the hills of the Rhône Valley and went to Paris to seek his fortune as a playwright. A professor of literature at the Sorbonne read one of his plays, Arthur of Brittany, and counseled Bernard to enroll in medical school instead. Heeding this advice, Bernard entered the Collège de France in the fall (Bernard 1979).

There he became intrigued by lectures in physiology given by François Magendie. Most chemists and physiologists of the time believed that only plants synthesized lipids, carbohydrates and proteins, whereas animals merely degraded them. The macromolecules within the body were therefore assumed to come largely preformed from the diet (Holmes 1974). A few skeptics questioned these ideas, because there sometimes seemed to be more fat in an animal's body than could have come from its diet. Bernard was captivated by Magendie's demonstrations of the intricacies of animal physiology, and from 1841 to 1844 he served as his laboratory assistant, gaining knowledge of techniques in animal experimentation (Holmes 1974).

Studies on Glucose

Soon Bernard began his own experiments, studying digestion and certain functions of the nervous system. He extended the digestion studies to examine the fate of sugars within the body and demonstrated that cane sugar (sucrose) was converted to grape sugar (glucose) in the gastrointestinal tract (Grmek 1968). Cane sugar, injected directly into a vein, was excreted unchanged in the urine; injected grape sugar, however, disappeared. Thus, glucose seemed to be the major form of sugar used within the animal body, and when Magendie, assisted by Bernard, fed starch to a dog, glucose was found in the dog's blood. Thus, glucose was a normal constituent of blood, at least after starch consumption, not just a sign of the diabetic condition as had been thought previously (Grmek 1968).

Early in 1848, Bernard began a systematic study to learn where glucose is used within the body. Following Lavoisier's idea, he conducted experiments with dogs that he thought would show glucose was burned in the lungs; however, these experiments yielded contradictory or uninterpretable results (Grmek 1968, Holmes 1974). In the early experiments, he had only insensitive methods for detecting and quantifying glucose and needed to use large quantities. He was assuming that these large quantities would be used almost instantly. Moreover, he used animals of various physiological conditions, and he sometimes fed the glucose and sometimes injected it. Improvement of a method for the detection of glucose based on its ability to reduce copper in an alkaline potassium tartrate solution significantly improved his results. Gradually Bernard improved his experimental techniques, and the early work set the stage for later, more successful, experiments.

The Source of Blood Glucose

In July 1848, Bernard conducted an experiment with a female that had been nursing a litter of pups. He did not feed her for one day and, as expected, found no glucose in her gastrointestinal tract, but to his surprise, he did find glucose in her blood (Grmek 1967). "What was the source of this glucose?" After this experiment, he altered the direction of his research to find the answer to this question (Grmek 1968, Holmes 1974).

In August, Bernard used a dog that had been fed only meat (no carbohydrates) for eight days and found a large amount of glucose in the portal vein, smaller amounts in the heart and the neck, but none in the chyle, stomach, intestine or urine. He exclaimed that the source of this glucose was "incomprehensible" (Grmek 1968).

A few days later, using a dog that had been fed only lard and tripe, he found no glucose in the mesentery (before the portal vein), but "enormous" quantities of glucose in the liver (Grmek 1968). Within the next four days, Bernard measured the glucose content of the liver of many different species, finding significant amounts of glucose in most. He concluded that liver of healthy animals contains glucose independent of a source of glucose in the diet (Bernard 1850, Bernard and Barreswil 1848).

Subsequent experiments provided evidence that the liver was the source of glucose in the body. By placing a tie between the liver and the portal vein, Bernard was able to show that the source of glucose previously found in the portal vein was the back flow of blood from the liver, not an alternative source prior to the liver (Bernard 1849 and 1850). For this work he received the Prize for Physiology in 1851 (Olmsted 1938) and the doctorate of science (Bernard 1853). It was also the beginning of Bernard's important concept of the body's ability to regulate its internal environment (Bernard 1878).

Search for the Source of Glucose in the Liver

In 1855 Magendie died, and Bernard was named to the Chair in Physiology at the Collège de France (Olmsted 1938). In this role he continued to pursue a variety of studies related to digestion, diabetes, toxins and the nervous system. During this time, he typically measured glucose in duplicate in several tissues of the animals he was studying. On one occasion he made the first measurement on a liver on one day, but did not make the duplicate measurement until the following day, after the liver had been allowed to stand at room temperature overnight. To his surprise, the content of glucose had increased markedly (Bernard 1855, Grmek 1967).

Bernard next decided to perfuse an isolated liver with cold water until the perfusate was free of glucose. He then allowed the liver to stand at room temperature for some hours. Upon resuming perfusion, he again found significant amounts of glucose in the perfusate. It appeared, therefore, that something within the liver was giving rise to glucose and it clearly was not making glucose from other elements in the blood. He took this as proof that there was a source of glucose within the liver (Bernard 1855).

Bernard then began the tedious job of isolating the glucogenic material present in the liver. He eventually recognized its chemical similarity to starch in plants and reported that it was present in opalescent extracts of the liver and formed a white precipitate when alcohol was added. It gave a red-wine color with iodine and was hydrolyzed by diastase, from saliva or plants, to produce glucose (Bernard 1857a, 1857b, and 1857c).

A Productive Decade

Within 10 years, Claude Bernard had made three major discoveries: 1) Glucose is a normal constituent of liver. 2) Liver is the source of blood glucose. 3) Liver forms glucose and stores it as glycogen, which, upon degradation, yields glucose.

Bernard's experiments and the theories he derived from them were major contributions to the science of nutritional physiology. His exceptional skill in the surgery required for these studies, and the understanding that he developed regarding the use of intact animals in experimentation, earned him recognition as the "father of experimental medicine." His major textbook (Bernard 1865) became a classic in the field, and he later received many honors, including membership in L'Académie Française (Bernard 1979, Olmsted 1938).

Literature Cited


Paper 2: Protein Cannot Be the Sole Source of Muscular Energy (Fick, Wislicenus and Frankland, 1866)

Presented by Kenneth J. Carpenter, University of California, Berkeley, CA 94720-3104 as part of the minisymposium "Experiments That Changed Nutritional Thinking" given at Experimental Biology 96, April 16, 1996, in Washington, D.C.

By 1865 it had been the general "textbook" view for over 20 years that the energy needed for muscular contraction came from the destruction of a portion of the muscle's own substance, i.e., protein. This had been stated by the organic chemist Justus Liebig in his influential Animal Chemistry. On page 233, he added that the protein broke up during the release of energy and that the nitrogenous fraction was converted to urea and excreted by the kidney, so that the total amount of work performed (i.e., both internally, as in the heart muscles, and externally) was proportional to the nitrogen excreted in the urine (Liebig 1840).

Liebig's second point was essentially disproved by the finding that prisoners receiving a constant daily ration of food excreted no more urinary nitrogen during 24 h in which they had worked a treadmill than on days when they had rested (Smith 1862). However, it was still possible that protein had been the sole muscle fuel and that more had broken down on rest days by some alternative mechanism. It certainly seemed that nitrogen intake was the main determinant of its output.

A Swiss physiologist, Adolf Fick, saw that the best conditions for a critical experiment would be to do a considerable amount of measurable work while eating a protein-free diet. Then if the heat energy obtained from the oxidation of protein to urea and carbon dioxide were known, and also the relation of heat energy to mechanical work, it should be possible to determine whether the amount of body protein metabolized was sufficient to have powered the work done.

Fortunately, Fick's brother-in-law, the chemist Edward Frankland, was at work in England developing a method for measuring the heat of oxidation of organic materials. High pressure "bomb" calorimeters had not yet been developed, but he was able to ignite a mix of potassium chlorate and manganese dioxide with the test material in a little "diving bell" immersed in an insulated water tank. Using a series of controls to adjust his results for the rise in temperature of the bath, he was able to obtain an impressive set of results for a long series of food materials and for urea (Frankland 1866).

The most directly relevant results are set out in Table 1. He assumed that metabolized protein yielded one-third of its own weight of urea, and he therefore subtracted the residual gross energy of this quantity of urea when estimating the energy released from the metabolism of 1 g of protein in the body.

Table 1. Frankland's presentation of his results for the energy values of protein and urea1

[View Table]

By this time it also seemed well established that the mechanical equivalent of heat was such that the energy needed to raise 1 kg a distance of 423 m was at least approximately equivalent to 1 kilocalorie (Joule 1843).

Now the human trial was needed. Fick and his university colleague Johannes Wislicenus passed a night at a hotel near the foot of a convenient mountain. At 0500 h next morning they set out, carrying urine collection equipment, and walked steadily up a steep path until, at 1320 h, they reached another hotel at the summit. They were in a cold mist throughout the climb and did not believe that they had had significant losses from sweating. From noon the previous day until 1900 h on their exercise day, their only food was cakes made from starch paste fried in fat; they also drank strongly sweetened tea and some beer and wine over the period.

Their results for their urinary nitrogen excretion, and the subsequent calculations, with slight "rounding off" of their values, are set out in Table 2. It is seen that even without making any allowance for the internal work of breathing and respiration, and even if the muscular system were 100% efficient, the quantity of protein metabolized was insufficient to have provided the energy needed for their climb; in fact it was 51% for one subject and 43% for the other.

Table 2. The results from the climbing trial

[View Table]

The climbers concluded that "the burning of protein cannot be the only source of muscular power" (Fick and Wislicenus 1866). And Frankland, on page 684 of his review of these and other results, added: "Like every other part of the body the muscles are constantly being renewed; but this renewal is not perceptibly more rapid during great muscular activity than during comparative quiescence. After the supply of sufficient albuminized matter [protein] in the food to provide for the necessary renewal of the tissues, the best materials for the production, both of internal and external work, are non-nitrogenous material..." (Frankland 1866).

These conclusions were not immediately accepted, but they stimulated further long-term trials that were confirmatory, although Liebig himself never admitted in so many word that he had been wrong (Carpenter 1994).

Literature Cited


Paper 3: Inorganic Iron Can Be Used to Build Hemoglobin (Stockman, 1893)

Presented by Richard A. Ahrens, Department of Nutrition and Food Science, College of Agriculture and Natural Resources, University of Maryland, College Park, MD 20742-7521 as part of the minisymposium "Experiments That Changed Nutritional Thinking" given at Experimental Biology 96, April 16, 1996, in Washington, DC.

The condition of anemia was originally named morbus virgineus by Johannes Lange (Lange 1554). Lange was a physician of Lemberg and Rector of Leipzig University. He considered this disease to be peculiar to virgins and to be due to a retention of menstrual blood. His therapy involved instructing virgins afflicted with this disease to marry as soon as possible. He cited no less an authority than Hippocrates, in his treatise De Morbis Virginum, as also recommending marriage to cure this disease.

J. Varandal renamed this disease "chlorosis" (Varandal 1615). The popular English term was the "green sickness," referring to the greenish hue assumed by Caucasians when their blood is low in hemoglobin. Chlorosis soon became a central feature in medical textbooks describing the diseases of women. Because chlorosis was a sign of virginity, European artists often painted young women during this era with a greenish hue. In art, if not in fact, chlorosis was a widespread condition.

By the mid 19th century, the disease of chlorosis was accepted by many physicians as being associated with neurotic and hysterical manifestations (Bullough and Voght 1973). Chlorosis became a form of neurosis. This view of chlorosis was an impediment to the acceptance of dietary therapy for its treatment. It was a refinement of the view that anemia was due to virginity in women, but it continued to perpetuate a sex bias. Bullough and Voght (1973) pointed out that sex bias flourished during the latter half of the 19th century as a male "backlash" against women's demands for more education, greater political equality and the elimination of male stereotypes about woman's place. Medical practitioners were almost all men, and many of them were hostile to any change in the status quo in male-female relationships. Medical schools that had admitted a few women early in the 19th century began to reject female applicants purely on the basis of their sex. Nursing schools were established in growing numbers to provide a alternative for females. By the latter part of the 19th century, chlorosis became an extremely common diagnosis (Clark 1887). It is necessary to appreciate this historical context to understand some of the resistance to accepting a nutrient deficiency as the cause of this disease.

Pierre Blaud in France recommended the use of pills containing ferrous sulfate for the treatment of chlorosis (Blaud 1832). The average dose amounted to approximately 150 mg/d, and considerable success was achieved. Despite this success, however, there was considerable resistance to the acceptance of chlorosis as a simple dietary iron deficiency. One of the obstacles to be overcome was the just-discussed sex bias that tended to associate chlorosis with the neuroses of women. Another obstacle to be overcome, however, was the inability of investigators using the balance method to demonstrate that inorganic iron could be absorbed from the gastrointestinal tract. V. Kletzinsky conducted a series of experiments (Kletzinsky 1854). In all of his studies, the amount of iron recovered in the feces was almost exactly equal to the amount of inorganic iron ingested. A third obstacle to be overcome was the toxic effect of intravenous injections of ferrous sulfate in dogs.

During the 1880s, Gustav von Bunge wrote two influential papers in which he concluded that only organic sources of iron were of value in treating chlorosis (Bunge 1885 and 1889). Von Bunge's interest in iron dated back to 1874 when he analyzed both blood and milk and recognized that blood was rich in iron and milk had very little. He developed a philosophy that people are always best served when they get essential nutrients from foods. That philosophy also applied to iron. To quote von Bunge, "Why should a patient buy his iron in the pharmacy and not on the market with the usual foodstuffs?" This is a philosophy with many adherents today. As von Bunge implemented what he believed, however, it soon became a personal crusade in which he claimed that "the iron which the doctors give to chlorotics to form hemoglobin is not absorbed at all."

As Gustav von Bunge got into his crusade he was soon claiming that iron therapy was successful because of the power of suggestion. It was well known, after all, that most chlorotics were women and often exhibited nervous or psychic disturbances. He felt that this made them "highly suggestible." The true villains, according to von Bunge, were those who advised young women to practice vegetarianism. He spent much of what remained of his life arguing against vegetarianism and was enthusiastic about the nutritional value of meat in the human diet. He died in 1920, just as Prohibition was beginning as the "noble experiment" in the United States. He was an implacable foe of alcohol consumption all his life and looked forward to the results of this experiment with U.S. citizens as the guinea pigs. He anticipated that a model U.S. society would result from Prohibition and that Europe would then soon follow this great example (McCay 1953). It is undoubtedly fortunate that he did not live to see the result of this particular experiment.

When he wasn't blaming the power of suggestion for the beneficial effects of inorganic dietary iron on chlorosis, von Bunge had a second explanation. Bullough and Voght (1973) noted that such contradictions were common among researchers studying "women's diseases" during the late 19th century. Von Bunge adopted Kletzinsky's theory (1854) that susceptible patients became chlorotic through the production by gut bacteria of hydrogen sulfide, which then reacted with organic iron compounds in the ingesta to produce insoluble ferrous sulfide. If inorganic salts of metals having insoluble sulfides were given as dietary supplements in large quantities, these should take up most of the hydrogen sulfide, leaving more of the organic iron compounds free for absorption.


Fig. 1. Hemoglobin responses of chlorotic patients to ferrous sulfide (in keratin capsules, 550 mg/d), iron citrate (subcutaneous, 32 mg/d) and bismuth oxide (9.6 g). The response times between the initial and final observations were 12 d for ferrous sulfide, 10 d for iron citrate and 9 d for bismuth oxide. The percentages given on the y-axis are based on the clinical standard for hemoglobin in use in 1893. Generated from the data of Stockman (1893).
[View Larger Version of this Image (36K GIF file)]

Ralph Stockman of the Edinburgh University School of Medicine put Kletzinsky's, and thereby von Bunge's, theory to the test (Stockman 1893). He did tests on chlorotic patients to determine if inorganic iron worked directly or by the indirect mechanism of binding with hydrogen sulfide. His results are summarized in Figure 1. He gave subcutaneous daily injections of ferrous citrate providing 32 mg of iron to three chlorotic young women and found an increase from 44% to 52% of normal hemoglobin concentration in 10 d. After 24 d the women had blood hemoglobin concentrations that were 72% of normal. Stockman then tried giving another four subjects 550 mg/d of iron by mouth in the form of ferrous sulfide and enclosed in keratin capsules that released the iron salt in the small intestine. Iron in this form could not be expected to bind any additional hydrogen sulfide. Nevertheless he found an increase from 48% to 60% of normal hemoglobin concentration in 12 d. After 33 d the women had blood hemoglobin concentrations that were 84% of normal. He also gave 9.6 g/d of bismuth dioxide to chlorotic women having blood hemoglobin concentrations that were 55% of normal and found these hemoglobin levels to be only 54% of normal 9 d later. Manganese dioxide gave a similar result. These latter two salts were quite capable of removing hydrogen sulfide from the gut, but they had no value in treating chlorosis.

It would seem that the elegant refutation of Gustav von Bunge's hypothesis by Ralph Stockman in 1893 should have made it apparent that inorganic iron had great value as a nutrient. In another paper two years later, Stockman (1895) showed that chlorosis in young women was explained by their low overall intake of food, particularly of meat, which resulted necessarily in low iron intake, at a time when the combined burdens of growth and menstrual blood loss increased their need. He showed, through the use of a more specific analytical procedure for iron in foods that avoided interference from starch, that the diet of anemic young women was particularly low in iron, partly because these young women were eating so little, and most of that was bread.

The reputation of Gustav von Bunge at that time, however, far exceeded the reputation of Ralph Stockman. As Carpenter (1990) has pointed out, poorly conducted research continued to question the therapeutic value of inorganic iron in anemia through the 1920s. Gustav von Bunge died in 1920. The old concept of "chlorosis" is also long gone (Fowler 1936). However, precautions are still needed to ensure an adequate intake of iron. In the United States, white bread and many breakfast cereals are routinely fortified with inorganic iron, and pregnant women are advised to take iron supplements. In the Third World, particularly where hookworm infestation is a chronic drain on people's blood supply, iron deficiency anemia remains a serious problem.

Literature Cited


Paper 4: A Micronutrient Deficiency in Chickens (Grijns, 1896-1901).

Presented by Barbara Sutherland, Department of Nutritional Sciences, University of California, Berkeley, CA 94720-3104 as part of the minisymposium "Experiments That Changed Nutritional Thinking" given at Experimental Biology 95, April 11, 1995, in Atlanta, GA.

Grijns was aware that it was not just red rice that prevented polyneuritis but all unpolished rice, and he decided to continue to study the whole silver skin and not just to focus on the pigment. His first feeding experiments confirmed Eijkman's conclusions that polyneuritis was not caused by a lack of fat, protein or mineral (Table 2). In his 1901 report, Grijns remarked: "In judging the suitability of a food, we have not finished when we have determined the quantity of albumen ... fat, carbohydrates and salts, even when we have applied the corrections for digestibility. We can indeed calculate from this whether a balance of nitrogen will be possible with it and whether the work which must be performed bother internally and externally, can be obtained from it, but not whether permanent health is possible."

Table 2. Grijns' key experiments1

[View Table]

Grijns believed that a number of substances existed, whose actions were not explained, but which played an important part in the prevention of disease. He illustrated this idea with two examples: "how very difficult it is, in spite of all the chemical analyses of mother's milk, to find a good substitute for it and how frequently we find that, when we think one has been found, we are again disappointed" and "the peculiar fact that scurvy, which usually develops from lack of fresh food, which sometimes occurs on long sea voyages, is usually cured when the patients can again obtain fresh meat and fresh greens." He concluded that still-unknown substances may be responsible.

Grijns used two approaches for investigating these "unknown substances." One was to prepare different fractions from the silver skin, and the other was by comparative assay (Grijns 1901). He first boiled rice bran in a large quantity of water for 24 h and then strained, filtered and evaporated the liquor to give a dried extract. He used fowls that were already consuming a polished rice diet and gave them the extract via a stomach tube. All the birds died with symptoms of polyneuritis. Increasing the dose of the extract further had no effect; neither did feeding the residue from the extracted bran. Grijns concluded that the "protective substances of the silver skin were for the most part lost through the methods of preparation used."

Grijns was also looking for a food material that when given in small amounts with polished rice, would prevent an outbreak of polyneuritis. He tested the mung bean (which he had noticed was often included in chicken feed) and the soybean. The results of his feeding experiments showed that both the skin and kernel of the mung bean prevented polyneuritis; however, the soybean was less effective. Comparing the composition of these two legumes, he saw that soybean was the richer in protein, fat and minerals but less effective as an anti-neuritic substance (Table 3). This supported his belief that polyneuritis was not caused by a lack of these three nutrients. In later experiments he found that extracts of mung bean were just as labile as those from silver skin. He stated that "we therefore had the same experience with Phaseolus radiatus (mung bean) as with the seed coat of the rice ... at every attempt to isolate the active constituents, they perished ... in different conditions they apparently became decomposed" (Grijns 1901).

Table 3. Composition of mung bean (P. radiatus java) and soybean (S. hispida tumida java)1

[View Table]

Eijkman had reported that the addition of some meat to sago, tapioca and arenga starch diets did not prevent polyneuritis. However, removing starch and feeding meat alone did cure the condition. From these results, Eijkman had concluded that starch was a significant harmful factor in the etiology of polyneuritis, but this explanation did not satisfy Grijns. He felt it important to determine whether polyneuritis could develop independently of starch consumption. He therefore fed four birds meat that had been extracted with water for 2 d, and all died with signs of polyneuritis. He then fed eight birds meat that had been autoclaved, and six of these also developed polyneuritis. Thus Grijns concluded that the development of polyneuritis was not connected with starch and was even wholly independent of the presence of carbohydrate. These experiments also confirmed that the nerve degeneration was not caused by a lack of protein (Table 2).

In a discussion of polyneuritis and beriberi, Grijns put forth two explanations for the symptoms that occurred: "either we presume a deficiency, a partial starvation, ... or ... there is a microorganism which exercises a degenerative influence on the nerves" (Grijns 1901). Concerning the possibility of a deficiency or partial starvation, Grijns stated that very little was known about the metabolism of the peripheral nervous system and that "if for the maintenance of the peripheral nervous system, a certain substance or group of substances is indispensable, which are immaterial for the metabolism of the muscles, then it may be assumed that very little of them is necessary. When therefore in certain foods the substances indispensable for the nervous system are lacking or are present in insufficient quantity, in the first place any reserve supply, which is present either in the nerve itself or in the blood or in some other organ, will be used up ... (and) disturbances will develop."

He explained that polyneuritis did not develop with total starvation, because in this situation the muscles were drawn upon to provide the needed protein and that this process released the "protective substance," which therefore became available to the nerves so that degeneration was prevented. Grijns used the notion of individual differences to account for why some birds did not develop polyneuritis: "one person needs a much larger quantity of food than another to maintain his physical equilibrium, while doing the same work ... If therefore the total metabolism shows important differences, there is no reason why, separate tissues which together furnish the total metabolism, should not have individual differences. Therefore a food which contains just enough of the still unknown nerve nutritive substances for one fowl contains too little for another."

In regard to the concept of a microorganism causing nerve degeneration, Grijns believed that this depended on the nourishment of the tissue to resist infectious organisms. He concluded that, irrespective of the causal factor of polyneuritis, "there occur in various natural foods substances which cannot be absent without serious injury to the peripheral nervous system ... The distribution of these substances in the different food stuffs is very unequal ... Attempts to separate them have resulted in their disintegration ... (showing) they are very complex" (Grijns 1901).

Recently we reported on Christiaan Eijkman's work on polyneuritis in chickens performed during the 1890s in Indonesia (Carpenter and Sutherland 1995). The timeline shows how early it was when this work was being performed: at the same time as Atwater's calorimetry work, and before the "vitamine" concept of Funk and McCollum's studies of fat-soluble factors (Table 1). This was a time when the infectious theory of disease was dominant, and it was while Eijkman was looking for an infectious cause of beriberi, a serious disease in Indonesia at that time, that he recognized a similarity with polyneuritis seen in chickens (Eijkman 1990). He found that polyneuritis appeared in fowls when they were fed a diet of polished rice, but that by adding the silver skin, which had been removed during polishing, the polyneuritis could be prevented or cured. From the results of many feeding experiments, Eijkman concluded that the polyneuritis was due to a nerve poison produced during the fermentation of starch in the chicken's crop and that the silver skin contained an antidote to this poison.

Table 1. Dates of some significant papers in nutritional science

[View Table]

Eijkman's work was cut short by ill health, and in 1896 he had to return to Holland. The research was continued by another young Dutch military surgeon, Gerrit Grijns. He had obtained his medical education in Holland and then studied physiology in the laboratory of Carl Ludwig in Germany (Grijns 1901). In 1892 he was sent to Indonesia to assist in another of Eijkman's studies, that of the physiological adaption of Europeans to tropical conditions. But Grijns was shortly recalled to military service, and when he was able to return to Batavia (modern-day Jakarta), Eijkman had already left for Holland. Grijns was then appointed to carry on the investigations into the cause of polyneuritis in chickens.

His official commission was to "investigate the physiological and pharmacological properties of the tannin contained in red rice" and to determine if the pigment found in red rice could be considered as a curative or preventive remedy for beriberi.

Literature Cited


Paper 5: Dietary Protein Standards Can Be Halved (Chittenden, 1904)

Presented by Vernon R. Young and Yong-Ming Yu, School of Science and Clinical Research Center, Massachusetts Institute of Technology, Cambridge, MA 02139 as part of the minisymposium "Experiments That Changed Nutritional Thinking" given at Experimental Biology 96, April 16, 1996, in Washington, DC.

The essentiality of a dietary substance, which was later named "protein" by the brilliant Swedish chemist Jac Berzelius (Korpes 1970), had been recognized by the middle of the 18th century by Beccari and by Haller (Munro 1969 and 1985). However, it was not until about a century later that definitive pronouncements were made about the dietary needs for proteins in human subjects. Thus, surveys of diets in the United Kingdom by Lyon Playfair, in Germany by Carl Voit, in the United States by Wilbur Atwater, as well as by others in other countries, revealed, in relation to protein intake and the total fuel value of the diet, that "all over the world people who can obtain such food as they desire use liberal rather than small quantities ..." (Benedict 1906). It was from these kinds of data that conclusions were drawn about the necessary intakes of protein, and Voit, who commanded considerable attention and scientific respect, concluded that---based on his assessment of his work in Munich---the protein intake of the average working man should be 118 g daily and that higher intakes would be necessary for heavy workers. Atwater, a pupil of Voit, supported this conclusion (Table 1). However, and in part through the expanded use of the nitrogen balance approach (developed initially by Boussingault for his studies in Alsace on the utilization of foodstuffs by milch cows), others began to question whether intakes lower than those shown in Table 1 would not only be adequate but possibly offer benefits for improvements in health. Arguably, the most significant of these others was Russell Henry Chittenden. Thus, Benedict (1906) says: "Of all the experiments heretofore made in which the low protein diet was used, none can compare with the exhaustive series of experiments recently completed by Professor Chittenden of New Haven." Indeed, they were impressive and after the "dust had really settled" they were destined to have an enduring and profound effect on the course of research in, and thinking about, human nutritional requirements.

Table 1. Some early dietary standards ("minimum for average man, under average conditions, doing moderate work, in health and strength")1

[View Table]

In 1882, Chittenden was appointed Professor of Physiological Chemistry at the Sheffield Scientific School at Yale University, two years after he had completed his Ph.D. degree in physiological chemistry, the first such degree awarded by a university in the United States (Vickery, 1945).

Chittenden (1904) presented a detailed account of his series of experiments in a monograph entitled Physiological Economy of Nutrition: With Special Reference to the Minimal Proteid Requirement of the Healthy Man. An Experimental Study. This is remarkable considering that his experiments began only in 1902 and continued well into 1904 and that this occurred well before the convenience afforded by computer-based data retrieval and summary techniques, not to mention desktop publishing. In this publication, he indicates that he had first questioned the premise that the dietetic standards adopted by mankind represented the real needs or requirements of the body (p. 3, Chittenden 1904): "We may even question whether simple observation of the kinds and amounts of foods consumed by different classes of people under different conditions of life have any very important bearing upon this question." He was the sort of mentor that any student would have been privileged to serve under: willing to challenge dogma and chart an entirely new experimental approach.

His experiments began with an opportunity to observe for several months the dietary habits of Horace Fletcher, an American of independent means. Chittenden noted that Fletcher's nitrogen intake averaged 7.19 g, and in the words of Dr. Anderson, the director of the Yale Gymnasium, "Mr. Fletcher of Venice performs this work with greater ease and with fewer noticeable bad results than any man of his age and condition I have worked with" (p. 14, Chittenden 1904).

Back then was no different than today, in that Chittenden needed financial support for the conduct of his investigations. He secured funds from the Carnegie Institution of Washington and the Bache Fund of the National Academy of Sciences, and he also received large donations from Fletcher and John H. Patterson of Dayton, Ohio.

The overall investigation consisted of three major experiments, each characterized by a long-term period of dietary protein restriction combined with nitrogen excretion measurements and supplemented in some studies with assessments of physical and mental well-being. Chittenden (1904) states "The writer, fully impressed with his responsibility in the conduct of an experiment of this kind, began with himself in November 1902." He therefore served as one of the subjects in his study of five university professors and instructors, including his former student Lafayette Mendel, who by that time had become professor.

On the basis of his own experience (Table 2), including the disappearance of the rheumatic problem he had been having in his knee joint and the findings with the other four Yale professionals, Chittenden concluded that the minimum "proteid" requirement was 93-103 mg N/kg body wt (about 0.6-0.64 g protein·kg-1·d), which anticipates, by 80 years, the mean requirement figure of 0.6 g protein·kg-1·d-1 proposed by FAO/WHO/UNU (1985)!

Table 2. A typical day's record of R. H. Chittenden's diet and nitrogen balance after 18 mo on his self-imposed experiment1

[View Table]

The next two series of studies confirmed and strengthened these initial findings; one of these was with 13 members of a detachment from the U.S. Army Hospital Corps, who were housed in Vanderbilt Square at Yale for 6 mo. The study included measures of physical and mental condition and of blood composition in addition to nitrogen excretion and balance over the 6-mo period. The conclusion was that 50 g of protein daily can establish nitrogen equilibrium and that there is, at this approximate intake, a maintenance of physical strength and vigor and an ability to respond to sensory stimuli. In a follow-up letter written to Chittenden by one of the participants, Private First Class J. Steltz, and on behalf of the other men, he stated: "The men are all in first-class condition. ... We eat very little meat now as a rule, and would willingly go on another test." The extensive findings in a third series involving eight Yale University athletes merely served to replicate all of these data and the interpretations that had been drawn from them.

Thus, Chittenden concluded that one-half of the 118 g of protein called for daily by the ordinary dietary standards is quite sufficient to meet all the real physiological needs of the body, certainly under ordinary conditions of life" (p. 475, Chittenden 1904).

Perhaps it ought to be noted that these experiments did not actually establish an average, minimum physiological requirement for dietary protein in these groups of subjects because 1) the low protein diets were freely chosen by the professional group, 2) the athletes were instructed to diminish the intake of protein but without imposition of a specific diet, and 3) the soldiers were given meals that contained lowered amounts of protein than provided by ordinary army rations and apparently with some reduction in the total "fuel value" of the food. The food given to each soldier was weighed, and at the close of every meal the uneaten food was determined and subtracted from the initial weight.

Although there has been some concern about the precision and accuracy of the nitrogen balance data generated from Chittenden's experiments (McCay 1912, Carpenter 1994), including the reliability of the urine and fecal collections and determination of nitrogen intake, the results of these series of experiments were coherent and dramatic. They presented a strong case that the physiological needs for protein were much lower than values represented by free-choice intakes of dietary protein.

Chittenden's conclusions were neither quickly nor universally accepted. For example, McCay (1912) referred to the onslaught to Chittenden's findings and ideas, but he used his own data from dietary surveys of Bengalis, as well as the data of others, to reach a conclusion that "Voit stands today absolutely vindicated." Although Cathcart (1911) was in "complete agreement with Professor Chittenden's statement that life can be maintained and frequently maintained at a high level on relatively low protein intake," he was not sure if it was desirable to keep a low intake as a general rule, and he expressed concern about the quality of protein. Later, he (Cathcart 1921) voiced reservations that were related to the lowered resistance to disease in persons consuming low protein diets. However, Chittenden (1911) had already argued that the problem with McCay's studies was that the diets of the populations studied in India lacked unidentified trace nutrients. This was probably true, in retrospect, given the public health problems of iron, vitamin A and iodine deficiencies that are prevalent in southeast Asia today.

The protein standards for healthy individuals continued to be set by the opinions of individuals until national and international committees were convened to establish dietary recommendations. An early international committee was set up by the League of Nations, and in 1936 the recommendation was that "the protein intake for all adults should not fall below 1 gramme of protein per kilogramme of body weight ..." (League of Nations 1936). No scientific justification was presented in support of the recommendation. In 1943 the U.S. Food and Nutrition Board of the National Academy of Sciences issued its first Recommended Dietary Allowances, and in this report 66 g of protein daily was recommended. These early figures proposed by expert groups were somewhat higher than those found to be sufficient by Chittenden, but they were far below the liberal standards that had been widely adopted during the middle 19th and on into the early part of the 20th century. It seems clear, however, that by the 1950s the metabolic approach used by Chittenden and others for establishing protein requirements had been well embraced, to the exclusion of the dietary intake approach followed by Voit, Atwater and others. For example, at the Princeton Conference in 1955, W. R. Aykroyd, the director of the nutrition division of the Food and Agriculture Organization, stated: "We need not linger on Carl Voit and his recommendations on protein requirements, which were over influenced by what was observed among the population of Munich in 1880" (FAO 1957a). Indeed, the first FAO report specifically concerned with protein requirements (FAO 1957b) depended heavily upon the review by Sherman et al. (1920) of the literature on nitrogen balance and adult requirements, which included extensive reference to Chittenden's work. Parenthetically, Sherman's paper might be viewed as a forerunner of modern-day meta-analysis! In any event, this 1955 FAO committee suggested that the average minimum requirement of adults for reference protein was 0.35 g/kg body wt and proposed a daily safe practical allowance of 0.66 g/kg. Although the more recent recommendations (FAO/WHO/UNU 1985) differ from those given in the 1957 report of FAO, this latter assessment undoubtedly would have given Chittenden great satisfaction, and it served as a vindication of the data obtained and conclusions drawn from his visionary studies, commenced 50 years earlier in the former New Haven residence of Joseph E. Sheffield.

Although Chittenden explored and made important contributions in the areas of digestive physiology and the action of proteolytic enzymes, an activity that had been enhanced by his year-long sojourn with Kühne in Heidelberg, and in toxicology, including heavy metal poisoning and disorders created by alcohol, it was his studies of the protein requirements of humans that may well be regarded as his greatest contribution to the advancement of nutritional science. When Chittenden died on Boxing Day (December 26) 1943, he had been a member of the National Academy of Sciences for more than 53 years!

Literature Cited


Paper 6: Liebig's Concept of Nutritional Adequacy Challenged (Hart et al., 1911)

Presented by Alfred E. Harper, University of Wisconsin, Madison, WI, as part of the minisymposium "Experiments That Changed Nutritional Thinking" given at Experimental Biology 96, April 16, 1996, in Washington, DC.

Imagine that we have fallen back 90 years through time. It is 1906. We know that protein and a few minerals (sodium, potassium, calcium, phosphorus, iron) are essential nutrients, but we are unaware of the essentiality of trace elements, vitamins or fatty acids. Liebig's concept from the 1850s that protein, a few minerals, and sources of energy (fat and carbohydrates) are the sole principles of a nutritionally adequate diet still dominates nutritional thinking and is widely accepted by leaders in the field, including Voit in Germany and Atwater and Langworthy at the U.S. Department of Agriculture (Harper 1993). Although several investigators have questioned the validity of Liebig's concept, their reports lie buried in the scientific literature (McCollum 1957, p. 201).

E. B. Hart has just been appointed Professor of Agricultural Chemistry at the University of Wisconsin to succeed S. M. Babcock. Babcock has observed that milk production by cows consuming rations composed of different feedstuffs differs considerably even when the rations are formulated to have the same gross composition (proximate analysis). He tells Hart that he is skeptical of Liebig's claim that the "physiological value" of a ration can be predicted from knowledge of its gross chemical composition (Hart 1932) and encourages Hart to test Liebig's hypothesis.

In 1907, in collaboration with G. C. Humphrey of the Animal Husbandry Department, Hart plans what will come to be known as the `Wisconsin single grain experiment'. He hires E. V. McCollum to conduct the chemical analyses and Harry Steenbock as a student assistant. The objective of the experiment is to compare the performance of four groups of heifers fed rations composed entirely of the corn, wheat or oat plant or a mixture of the three, all formulated to be closely similar in gross composition and energy content (Hart et al. 1911).

The animals, 16 Shorthorn heifers about 6 mo of age and weighing 300-400 pounds (lb), are to be carried to maturity and through two consecutive reproductive periods. The four rations, designed to provide 14 lb dry matter/d, consist of the following (lb/d): corn meal 5, corn gluten 2, corn stover 7; oat meal 7, oat straw 7; ground wheat 6.7, wheat gluten 0.3, wheat straw 7; and a mixed ration consisting of equal parts of the other three.

The average amounts consumed by the four groups during the course of the experiment (14.5 to 15.2 lb/d) did not differ appreciably. Values for crude digestibility of the different rations averaged 65 ± 3% for both dry matter and nitrogen and were not significantly different.

Weight gains of the groups after 1 and 3 y are shown in Table 1. Although the corn-fed group gained considerably more weight than the wheat-fed group, variability within groups was such that, as can be seen from the large SD, the results did not provide convincing evidence that the growth responses were different.

Table 1. Weight gain of heifers fed single grain rations1

[View Table]

The first clear evidence of differences in the responses of the groups was from observations on the appearance of the animals after the first year. Cows consuming the corn ration had smooth coats, were full through the chest, and appeared healthy. Those consuming the wheat ration had rough coats, were of smaller girth, and appeared gaunt. The other two groups were intermediate between the corn- and wheat-fed groups.

There were major differences in reproductive performance (Table 2). Calves born to cows consuming the corn ration were strong and vigorous in both years and all lived. The cows consuming the wheat ration all delivered 3-4 wk prematurely. Their calves were weak both years, and none lived beyond 12 d. Again, results for the other two groups were intermediate. In 1909, the first year of calving, cows consuming the oat and mixture rations produced weak calves, with only two from the oat group and one from the mixture group living beyond a few days. In 1910, the performance of these two groups was better. The calves were carried to term, and all of the calves from the oat-fed group and two from the mixture-fed group lived but were weaker than those from the corn-fed group. Average weights of the calves, were 78, 72, 62 and 49 lb for groups fed corn, oat, mixture and wheat, respectively.

Table 2. Condition and survival of calves1

[View Table]

Milk production of each group was measured for 30 d each year following cessation of colostrum secretion. The difference between the amounts of milk produced by the wheat- and corn-fed groups was large (Table 1). Average values for the groups (lb/d ± SD, number of observations in parentheses) were as follows: corn-fed group, 26 ± 3.5 (8); oat-fed group, 22.9 ± 6.5 (6); mixture-fed group, 20.4 ± 1.5 (5); wheat-fed group, 14.1 ± 2.6 (4). (In calculating the average for the wheat-fed group I deleted two values, one for a cow that died of illness, and one abnormally low value). No differences were observed in milk composition (total solids, total protein, casein, ash or fat) or in the characteristics of the milk fat.

The wheat ration was known from chemical analyses to provide less calcium, magnesium and potassium than the other rations. Two cows in the wheat-fed group were therefore given a supplement of these minerals during 1 y of the study to raise their levels of intake to those provided in the diets of the other groups. The condition of the cows was not noticeably improved. The calf produced by one of them was small and weak and lived only a few hours.

After the experiment was completed, some animals were switched to other rations for an additional year (1910-1911). The vigor and health of one cow that was switched from wheat to corn improved rapidly. It produced a calf weighing 81 lb, compared with 47 and 48 lb for calves produced the previous 2 y. One cow that was switched from the corn diet to the wheat ration supplemented with extra calcium, magnesium and potassium deteriorated. Its calf was stillborn 18 d prematurely and weighed only 36 lb, compared with 93 and 85 lb for those born the previous 2 y.

The authors concluded that 1) the nutritive value of a ration could not be predicted reliably from measurements of its total digestible nutrients and energy content, 2) the differences in performance were unlikely to be due to differences in the protein component because animals fed the diet that provided a mixture of proteins did not grow as well as some of those fed the single grain diets, and 3) mineral inadequacies were unlikely because the mineral supplement did not improve the performance of cows consuming the wheat ration. They did not claim to have eliminated these possible explanations conclusively.

They stated "we have no adequate explanation of our results." They did not attribute the inferior performance of the wheat-fed group to a deficit of some unidentified essential nutrient. They did, however, propose that the physiological value of a food or feed could be determined by measuring growth or other responses of animals fed diets in which a portion of a basic ration was replaced by the product to be tested.

Is it possible, with hindsight, to identify specific nutritional deficits that can account for the inferior performance of the cows fed the wheat ration? Levels of calcium and magnesium were low in the wheat ration. A level of 0.16% of magnesium is adequate for dairy cattle, but the level of 0.16% for calcium is marginal (Shepherd and Converse 1939) and 0.3% is now recommended (Scott 1986). Nonetheless, the reproductive performance of cows consuming the wheat ration, as Hart and colleagues noted, was not improved when they were given supplements of calcium and magnesium. Also, the fat content of the wheat ration was low. Dairy cattle require at least 2% of fat (Shepherd and Converse 1939). Low milk production, as was observed in the wheat-fed group, is an early sign of fatty acid deficiency in lactating dairy cows.

In addition, the carotenoid content of forage crops declines during storage, and vitamin A depletion occurs commonly in cows maintained during the winter on forage that has been stored for several months. A predominant sign of vitamin A deficiency in dairy cows is premature calving, with the calves often born dead or surviving for only a short time. Reproductive performance of cows receiving the corn ration, which would be expected to provide a high level of carotenoids, was excellent. That of the cows fed the wheat ration was poor. McCollum (1964) attributed this to loss of most of the leaves of the wheat plant during threshing. However, reproductive performance of the oat-fed group was poor the first year but much better the second, suggesting that the carotenoid content of feedstuffs varies from year to year. It would thus seem highly probable that differences in reproductive performance of the groups were due to differences in vitamin A status, owing to differences in the carotenoid content of the rations, possibly complicated in the case of the wheat-fed group by an inadequate intake of fat and a marginal intake of calcium.

The results of this experiment provided the first clear evidence from research in the United States that the nutritive value of a diet