Vitamin A, Agene Flour and Edward Mellanby's Legacy

Sir Edward Mellanby (1884–1955) is remembered first for the dog experiments that showed cod liver oil prevents rickets, the work that led to the discovery of vitamin D. But the rickets papers were only the beginning of a research life that ran for almost four decades and ended, by the accounts of his biographers, with him still at work on the day he died. In the late 1920s he turned to the other great vitamin of cod liver oil, vitamin A, and argued that it helped the body resist infection. In the 1930s and 1940s he followed vitamin A into the growing bones and nerves of young animals, and late in life he took the question into the tissue-culture dish with the cell biologist Honor Fell. And in 1946 his long study of how cereals affect dogs produced one of the most consequential food-safety findings of the century: a chemical used to bleach and “improve” flour was making dogs ill.

This page follows those later threads in turn — the “anti-infective” vitamin A idea and the interwar trials it inspired, the bone-and-nerve experiments, the agene flour story and the chemists who isolated its toxic factor, his concern about the chemical manipulation of food, and how vitamin D and vitamin A science developed after him. His early life, the rickets experiments themselves and the Mellanbys’ cereal and tooth research each have their own pages in this wing, linked under Connections.

Table of Contents

  1. 1. Vitamin A as an Anti-Infective Agent
  2. 2. Puerperal Sepsis Trials, 1929–1931
  3. 3. Interwar Vitamin A Trials and Their Limits
  4. 4. Vitamin A, Bone and Nerves in Young Dogs
  5. 5. Tissue Culture with Honor Fell
  6. 6. Canine Hysteria and Agene-Treated Flour (1946)
  7. 7. Isolating the Toxic Factor
  8. 8. The Chemical Manipulation of Food
  9. 9. Vitamin D Science After Mellanby
  10. 10. Later Vitamin A Research and Mellanby’s Place in Nutrition
  11. Key Research Papers
  12. Connections

1. Vitamin A as an Anti-Infective Agent

By the mid-1920s the “fat-soluble factor” that Mellanby had found in cod liver oil was known to be at least two substances. Elmer McCollum’s group in the United States had shown in 1922 that the calcium-depositing, rickets-preventing activity (soon called vitamin D) could be separated from the growth-promoting and eye-protecting activity that kept the older name, vitamin A. Mellanby had first believed that vitamin A itself prevented rickets; once the two were separated, he took up vitamin A as a research subject in its own right.

In October 1928 he and his Sheffield colleague H. N. Green published a paper in the British Medical Journal under a deliberately bold title: “Vitamin A as an anti-infective agent.” Drawing on observations in laboratory animals kept on diets lacking vitamin A, they argued that the vitamin did more than protect the eyes and support growth — that animals short of it were unusually prone to infection, and that the vitamin had a role in the body’s resistance to infection.

The phrase stuck. Richard Semba of the Johns Hopkins University School of Medicine, writing in the Journal of Nutrition in 1999, describes the idea that vitamin A could be used in “anti-infective” therapy as one that emerged in the 1920s and was “largely championed by Edward Mellanby.” It is the starting point of a line of research that, as the later sections show, ran on into the large child-health trials of the late twentieth century.

Why cod liver oil mattered here too

The natural source sits at the centre of this story just as it did with rickets. Cod liver oil was the richest everyday food source of both vitamins A and D then available, and it was the practical form in which “vitamin A therapy” reached patients in the interwar years. Semba notes that in the trials which followed, vitamin A was usually supplied as cod liver oil. That has a consequence worth keeping in mind: many interwar studies could not cleanly separate any effect of vitamin A from an effect of vitamin D or of the oil as a whole. The site’s page on cod liver oil describes the food itself.

Back to Table of Contents

2. Puerperal Sepsis Trials, 1929–1931

Puerperal sepsis — infection of the womb and bloodstream after childbirth, often called childbed fever — was still one of the leading causes of death among mothers in Britain in the late 1920s. There were no antibacterial drugs; doctors could offer hygiene, nursing and surgery, and little else once the infection had spread into the blood. It was a natural test of the idea that vitamin A strengthened resistance to infection.

In June 1929 Mellanby and Green published a second BMJ paper, “Vitamin A as an anti-infective agent: its use in the treatment of puerperal septicaemia,” reporting on the vitamin given to women who were already seriously ill with blood infection after childbirth. Two years later, in October 1931, Green, Mellanby and two colleagues, D. Pindar and G. Davis, turned from treatment to prevention in “Diet as a prophylactic agent against puerperal sepsis,” describing a diet enriched with vitamin-rich foods given to women before delivery.

These papers are the clearest example of Mellanby carrying a laboratory hypothesis straight into the clinic. They also show the methods of their time. As Semba’s review of the period explains, studies of this era generally lacked the features that are now standard in a controlled trial, so their results are best read as observations that encouraged further work rather than settled proof. The wider history of childbed fever, from Semmelweis onward, is told on the site’s Sepsis: History and Discovery page.

Back to Table of Contents

3. Interwar Vitamin A Trials and Their Limits

Mellanby’s papers helped launch two decades of clinical work. Semba counts at least 30 trials between about 1920 and 1940 that set out to learn whether vitamin A, usually in the form of cod liver oil, could reduce illness and death from respiratory disease, measles, puerperal sepsis and other infections.

What the trials lacked

Semba’s central point is methodological. The early studies generally lacked the tools that define the modern controlled trial: random allocation of patients to treatment or control, masking (so that neither patients nor assessors knew who received what), sample-size and statistical-power calculations, and placebo controls. Without them, it is hard to tell a real effect from chance, from differences between the groups, or from the hopes of the investigators.

Mixed results and advertising

The results, by Semba’s account, were mixed — some studies reported benefit, others did not. He also records that the pharmaceutical industry emphasised the positive results in advertising to the public, so that the popular picture of vitamin A as an infection-fighter ran ahead of what the trials as a whole had shown.

Why interest faded

Then the field changed. In the mid-1930s the first sulfonamide drugs arrived, beginning with the work of Gerhard Domagk, and for the first time doctors had a drug that acted directly against bacteria such as the streptococcus behind most childbed fever. Semba writes that with the advent of the sulfa antibiotics, scientific interest in vitamin A as “anti-infective” therapy waned. The question was not answered so much as set aside — to be reopened, with better methods, half a century later (section 10).

Back to Table of Contents

4. Vitamin A, Bone and Nerves in Young Dogs

While the clinical trials ran their course, Mellanby’s own laboratory work on vitamin A moved in an unexpected direction: from infection to the skeleton and the nervous system. He had long been struck by nervous signs in young animals on deficient diets, and in a series of papers in the Journal of Physiology he traced them to the bones.

Deafness by diet (1938)

In December 1938 he published “The experimental production of deafness in young animals by diet,” showing that a defective diet in early life could damage hearing — a striking result at a time when deafness was rarely thought of as a nutritional question.

Bone overgrowth pressing on nerves (1941 and 1943)

Two longer papers set out the mechanism he proposed. “Skeletal changes affecting the nervous system produced in young dogs by diets deficient in vitamin A” (1941) and “The effect of bone dysplasia (overgrowth) on cranial nerves in vitamin A-deficient animals” (1943) described how, without enough vitamin A, the growing bones of young dogs did not remodel normally. Bone that would usually be shaped and trimmed as the skeleton grew instead overgrew, narrowing the bony channels through which nerves pass, so that cranial nerves and parts of the nervous system were compressed and damaged.

The idea was important beyond dogs. It suggested that vitamin A is needed not only for the eye and for the linings of the body, but for the orderly shaping of bone during growth — a role that tissue-culture work would soon look at more directly. The site’s Vitamin A page describes the vitamin’s known functions.

Back to Table of Contents

5. Tissue Culture with Honor Fell

Whole-animal experiments have a limit: when a dog on a deficient diet shows bone changes, it is hard to say whether the vitamin acts on the bone itself or through some other organ. Tissue culture — keeping living pieces of tissue alive in a dish — offered a way around that. Its leading British practitioner in bone and cartilage was Honor Fell (1900–1986) of the Strangeways Research Laboratory in Cambridge, and in his last years Mellanby worked with her.

Their best-known joint paper, “The effect of hypervitaminosis A on embryonic limb bones cultivated in vitro,” appeared in the Journal of Physiology in 1952. Instead of a shortage of vitamin A, it examined the opposite: what happens to developing limb bones taken from embryos and grown in a culture medium containing an excess of the vitamin. Because the bones were growing on their own in the dish, any change could be put down to the vitamin acting directly on skeletal tissue. The collaboration continued with further papers on vitamin A and cultured tissues into the mid-1950s, according to the Royal College of Physicians’ record of his work.

The pairing is a neat summary of Mellanby’s career: he began with the crudest possible experiment — feeding puppies porridge — and ended with one of the most refined techniques of mid-century biology, still asking the same question about what a food factor does to growing bone.

Back to Table of Contents

6. Canine Hysteria and Agene-Treated Flour (1946)

The discovery for which Mellanby’s later career is best known came, as the nutrition historian Kenneth Carpenter puts it, as a spin-off of his continued work on how cereals affect the skeleton. Ever since the rickets experiments he had kept dogs on diets built around cereals and bread, and the problem he turned to was “canine hysteria,” also called running fits — a disorder that, Carpenter records, had developed in dogs in the 1930s and 1940s.

The agene process

Much white flour at the time was treated with nitrogen trichloride, sold under the name “agene.” Millers used it as a flour “improver”: a small dose of the gas was used to bleach the flour and change the baking behaviour of the dough. It was a clear example of a chemical process applied to a staple food on a very large scale.

The 1946 paper

In December 1946 Mellanby reported in the British Medical Journal, in a paper titled “Diet and canine hysteria; experimental production by treated flour,” that he could produce canine hysteria in dogs experimentally by feeding them flour that had been treated by the agene process. A disorder seen in dogs for more than a decade, in other words, could be traced to treated flour in their food. A further short report in August 1947 described more observations on the production of canine hysteria by flour treated with nitrogen trichloride.

Why the finding was unsettling

Carpenter draws out the troubling detail: agenised flour was nontoxic to rats. A safety test on the standard laboratory animal would have found nothing. Only because Mellanby happened to be feeding dogs large amounts of flour did the effect come to light. Carpenter records that, after the finding, the use of the process by millers was banned. (The dates of the withdrawal in Britain and the United States are not given in the sources used for this page.) The site’s Wheat page covers the grain itself.

Back to Table of Contents

7. Isolating the Toxic Factor

Mellanby’s result set chemists a clear task: find the substance in agene-treated flour that caused the fits. Two groups went after it.

The 1949–1950 identification

A team of British cereal chemists — H. R. Bentley, E. E. McDermott, J. Pace, J. K. Whitehead and T. Moran — published a short note in Nature in 1949 on the action of nitrogen trichloride on proteins, announcing the isolation of a crystalline toxic factor, followed in 1950 by a full paper in the Proceedings of the Royal Society. Its abstract sets out the chemistry. Proteins relatively rich in the amino acid methionine — zein from maize, gluten from wheat, and egg albumen — became toxic to certain animals when treated with nitrogen trichloride. The toxicity survived digestion of the proteins, and from each of the three the same pure crystalline substance was isolated. It was a derivative of methionine, with the molecular formula C5H12N2O3S.

Mellanby’s own isolation (1951)

Working in parallel, Mellanby with the biochemists P. N. Campbell and T. S. Work reported “The isolation of a toxic substance from agenized wheat flour” in the Biochemical Journal in January 1951 — taking the substance directly from the treated flour rather than from purified proteins.

Methionine sulphoximine

The compound became known as methionine sulphoximine (in American spelling, sulfoximine). Mellanby continued to study it: one of his last papers, published in 1956 after his death, reported “preliminary experiments on the effect of methionine sulphoximine on the developing chick and on transplantable tumours.” The same molecule later became a standard laboratory tool in biochemistry as an inhibitor of the enzyme glutamine synthetase, which is how it appears on the site’s page on endogenous glutamine synthesis.

Back to Table of Contents

8. The Chemical Manipulation of Food

The agene episode left Mellanby with a wider question. Here was a chemical treatment applied for decades to one of the most basic foods in the national diet, and its effect on an animal had come to light only by chance — and only in one species. In October 1951 he published a lecture in the British Medical Journal under the title “The chemical manipulation of food,” and the same title was used for a short book that year.

The title itself expresses the concern the agene work had raised for him: that the industrial processing and chemical treatment of staple foods deserved the same careful scientific scrutiny as a medicine. It connects naturally with the rest of his career. His rickets work had shown that what was missing from a diet mattered; his cereal work had shown that something present in an ordinary food (phytic acid) could interfere with a nutrient; and agene showed that something added to a food in processing could do harm that standard testing missed.

That line of questioning — what processing adds to, and takes away from, a food — continues in the site’s coverage of modern flour treatments such as azodicarbonamide.

Back to Table of Contents

9. Vitamin D Science After Mellanby

Mellanby did not discover vitamin D as a chemical; his 1919 dog experiments identified a fat-soluble dietary factor in cod liver oil that prevented rickets, and others separated, named and purified it. But the research that his experiments set going transformed medicine, and it kept developing long after his death.

From rickets cure to a hormone system

The biochemist Hector DeLuca, in his 1988 review “The vitamin D story,” summarises the arc. The discovery between 1919 and 1924 of vitamin D, and of its production in skin and foods by ultraviolet irradiation, led to the elimination of rickets as a major medical problem. The identification and chemical preparation of the vitamin in the following decade gave physicians large quantities of it for treating metabolic bone diseases.

Yet as late as the early 1960s, DeLuca notes, little was known about how vitamin D actually worked. Modern biochemistry then showed that vitamin D is first modified in the liver (25-hydroxylation) and then in the kidney (1-alpha-hydroxylation) to form the active vitamin D hormone, 1,25-dihydroxyvitamin D3. That process is tightly feedback-regulated and is one of the body’s major endocrine systems for controlling calcium and phosphorus in the blood and the state of the bones. Chemical synthesis of the hormone and its analogues opened treatments for conditions such as vitamin D-resistant rickets, hypoparathyroidism, kidney bone disease and osteoporosis, and DeLuca also describes work on psoriasis and on certain leukaemias.

The road from a puppy fed porridge and cod liver oil to a hormone made in the kidney is told from the vitamin’s side on the site’s Vitamin D3: History and Discovery page.

Back to Table of Contents

10. Later Vitamin A Research and Mellanby’s Place in Nutrition

The anti-infective question reopened

The question Mellanby raised in 1928 came back, with modern methods, in the 1980s and 1990s. Semba opens his history by noting that in the fifteen years before 1999 a large series of controlled clinical trials had shown that vitamin A supplementation reduces illness and death among children in developing countries, and he describes those trials as following a tradition of investigation that began largely in the 1920s.

The evidence has since been pooled. A 2022 Cochrane systematic review by Imdad and colleagues identified 47 studies involving roughly 1.2 million children aged six months to five years, in 19 countries. Across 19 trials reporting all-cause mortality, it found a 12% reduction in the risk of death for children given vitamin A supplements compared with controls (risk ratio 0.88; high-certainty evidence), and a 12% reduction in deaths from diarrhoea. The review found reductions in the incidence of diarrhoea, measles, night blindness and vitamin A deficiency, but no evidence of a difference in the incidence of respiratory disease or in deaths from measles or respiratory disease, and it reported an increased risk of vomiting in the first 48 hours after a dose. These are findings for children in settings where vitamin A deficiency is common; the site’s page on vitamin A deficiency and child mortality covers that background.

Seen against those results, Mellanby’s “anti-infective” label looks partly right and partly too broad: the modern evidence supports an effect on overall child survival and on some infections, but not on all of the conditions that the interwar enthusiasm covered, such as respiratory disease.

Mellanby’s place

Taken together, his later work shows a consistent habit of mind. He followed foods — cod liver oil, oatmeal, white flour — into the body, using the dog as his main experimental animal, and he took laboratory findings into public health: the rational treatment of rickets, attention to cereals and minerals, and the end of agene. Like his teacher Frederick Gowland Hopkins, he helped establish the idea that what is in — and what is done to — ordinary food shapes health. Some of his conclusions were later refined or corrected, notably his early belief that vitamin A prevented rickets and the broad claims of the anti-infective era. But the questions he posed about vitamins A and D, about cereals, and about the chemical treatment of food have remained live research subjects ever since.

Back to Table of Contents

Key Research Papers

  1. Green HN, Mellanby E. Vitamin A as an anti-infective agent. Br Med J. 1928;2(3537):691-6. PubMed PMID: 20774205
  2. Mellanby E, Green HN. Vitamin A as an anti-infective agent: its use in the treatment of puerperal septicaemia. Br Med J. 1929;1(3569):984-6. PubMed PMID: 20774721
  3. Green HN, Pindar D, Davis G, Mellanby E. Diet as a prophylactic agent against puerperal sepsis. Br Med J. 1931;2(3691):595-8. PubMed PMID: 20776417
  4. Semba RD. Vitamin A as “anti-infective” therapy, 1920-1940. J Nutr. 1999;129(4):783-91. PubMed PMID: 10203551
  5. Mellanby E. The experimental production of deafness in young animals by diet. J Physiol. 1938;94(3):380-98. PubMed PMID: 16995052
  6. Mellanby E. Skeletal changes affecting the nervous system produced in young dogs by diets deficient in vitamin A. J Physiol. 1941;99(4):467-86. PubMed PMID: 16995266
  7. Mellanby E. The effect of bone dysplasia (overgrowth) on cranial nerves in vitamin A-deficient animals. J Physiol. 1943;101(4):408-31. PubMed PMID: 16991575
  8. Fell HB, Mellanby E. The effect of hypervitaminosis A on embryonic limb bones cultivated in vitro. J Physiol. 1952;116(3):320-49. PubMed PMID: 14939182
  9. Mellanby E. Diet and canine hysteria; experimental production by treated flour. Br Med J. 1946;2(4484):885-7. PubMed PMID: 20278492 (the PubMed record spells the first word “Dist”)
  10. Mellanby E. Further observations on the production of canine hysteria by flour treated with nitrogen trichloride (agene process). Br Med J. 1947;2(4520):288. PubMed PMID: 20257560
  11. Bentley HR, McDermott EE, Pace J, Whitehead JK, Moran T. Action of nitrogen trichloride (‘agene’) on proteins: isolation of crystalline toxic factor. Nature. 1949;164(4167):438-439. DOI: 10.1038/164438a0
  12. Bentley HR, McDermott EE, Moran T, Pace J, Whitehead JK. Action of nitrogen trichloride on certain proteins I. Isolation and identification of the toxic factor. Proc R Soc Lond B. 1950;137(888):402-417. DOI: 10.1098/rspb.1950.0049
  13. Campbell PN, Work TS, Mellanby E. The isolation of a toxic substance from agenized wheat flour. Biochem J. 1951;48(1):106-13. PubMed PMID: 14820791
  14. Mellanby E. Preliminary experiments on the effect of methionine sulphoximine on the developing chick and on transplantable tumours. Br J Nutr. 1956;10(3):264-74. PubMed PMID: 13355937
  15. Carpenter KJ. Contribution of the dog to the science of nutrition. J Nutr. 1991;121(11 Suppl):S1-7. PubMed PMID: 1941202
  16. Mellanby E. The chemical manipulation of food. Br Med J. 1951;2(4736):863-9. PubMed PMID: 14869746
  17. McCollum EV, Simmonds N, Becker JE, Shipley PG. Studies on experimental rickets. Journal of Biological Chemistry. 1922;53(2):293-312. DOI: 10.1016/s0021-9258(18)85783-0
  18. DeLuca HF. The vitamin D story: a collaborative effort of basic science and clinical medicine. FASEB J. 1988;2(3):224-36. PubMed PMID: 3280376
  19. Imdad A, Mayo-Wilson E, Haykal MR, Regan A, Sidhu J, Smith A, Bhutta ZA. Vitamin A supplementation for preventing morbidity and mortality in children from six months to five years of age. Cochrane Database Syst Rev. 2022;3(3):CD008524. PubMed PMID: 35294044
  20. Hawgood BJ. Sir Edward Mellanby (1884-1955) GBE KCB FRCP FRS: nutrition scientist and medical research mandarin. J Med Biogr. 2010;18(3):150-7. PubMed PMID: 20798415

PubMed Topic Searches

  1. Vitamin A supplementation and child mortality
  2. Agene and canine hysteria
  3. Methionine sulfoximine and nitrogen trichloride
  4. Hypervitaminosis A and bone in culture
  5. Mellanby E [Author] and vitamin A

Further Reading

Back to Table of Contents

Connections