On this page
Rickets sits in every school list of deficiency diseases, next to the vitamin that prevents it. The list makes the link look obvious. A hundred years ago it was not. Doctors blamed slum air, lack of exercise, too little fat or infection. Edward Mellanby settled much of the argument with dogs, careful diets and a lot of bone analysis. This post explains what he did, works through his own data, and shows where he went wrong.
Who was Edward Mellanby?
Mellanby was a British medical scientist who lived from 1884 to 1955. When he gave his rickets lectures in 1919 he was Acting Superintendent of the Brown Institute, London University, and by his 1921 Medical Research Council report he was Professor of Pharmacology at the University of Sheffield. His Royal Society memoir records that he died suddenly on 30 January 1955, nearly six years after retiring as Secretary of the Medical Research Council, a post he held for 16 years. It also calls his wife, May Mellanby, his "devoted wife and scientific partner". She studied how rickets affects teeth.
How bad was rickets in 1919?
In his 1919 lectures to the Royal College of Surgeons, reprinted from The Lancet of 15 March 1919, Mellanby quoted earlier surveys. Schmorl's examination of children who died before the age of 4 found that 90% had had rickets. Lawson Dick, examining children in London County Council schools, put the figure at 80%. Mellanby also noted that hypotheses about the cause were "legion".
How did Mellanby test the causes?
He used young puppies, starting them on fixed diets at 5 to 8 weeks old. By 1919 the study, done for the Medical Research Committee, had used 200 puppies. He detected rickets three ways: X-rays of the bones, calcium measured in the bones after death, and microscope sections.
First he found a diet that reliably caused rickets: 175 cc of whole milk a day with porridge of oatmeal and rice, or later bread. Then he added one food at a time and watched what happened:
Raising whole milk from 175 cc to 500 cc a day prevented rickets.
Cod-liver oil, butter and suet prevented it well. Linseed oil did not.
Yeast had no protective effect, so the factor was not "water-soluble B".
Orange juice, enough to prevent scurvy, did not prevent rickets either.
Extra calcium phosphate did not help when the protective factor was missing.
He concluded that rickets "is a deficiency disease which develops in consequence of the absence of some accessory food factor", an "anti-rachitic factor" found mainly in animal fats. He thought it was probably the same as "fat-soluble A" or had "a somewhat similar distribution". The anti-rachitic vitamin soon got its own name: by 1928 Adolf Windaus, in the Nobel lecture quoted below, was calling it vitamin D.
Worked example: reading Mellanby's bone data
In a 1922 experiment described in his 1925 report, Mellanby gave puppies diets with different amounts of oatmeal. Some were exposed to a mercury-vapour lamp, which gives ultraviolet light. He measured calcium oxide in the shaft of each femur.
Puppy 712, normal oatmeal, no lamp: 0.68 g of calcium oxide.
Puppy 714, same diet, with the lamp: 1.14 g.
Puppy 713, double oatmeal, no lamp: 0.51 g.
Puppy 717, double oatmeal, with the lamp: 0.28 g more than puppy 713.
A normal littermate given cod-liver oil: 1.53 g.
Step 1: absolute gains. For normal oatmeal the lamp added 1.14 − 0.68 = 0.46 g. For double oatmeal it added 0.28 g, so 717 had 0.51 + 0.28 = 0.79 g.
Step 2: relative gains. 0.46 ÷ 0.68 ≈ 0.68, an increase of about 68%. 0.28 ÷ 0.51 ≈ 0.55, about 55%.
Step 3: compare with the healthy dog. Even with the lamp, 717's 0.79 g is only about half of 1.53 g (0.79 ÷ 1.53 ≈ 0.52).
Mellanby's conclusion: the more oatmeal, the worse the rickets and the less the light helped. He argued that cereals contain something that works against calcification. In 1939 he and D. C. Harrison published Phytic acid and the rickets-producing action of cereals, naming the suspect.
What did Mellanby get wrong?
In 1919 he played down sunlight and exercise. "My own experience is that confinement will not produce any symptoms of rickets in adequately fed puppies," he said. Around the same time, others were showing that ultraviolet light cures rickets. Adolf Windaus, in his 1928 Nobel lecture, credits the German physician Huldschinsky as the first to find that ultraviolet irradiation is a remedy, and Hess and Steenbock with showing that irradiating food works too.
By 1925 Mellanby accepted that sunlight could partly replace the vitamin in food. But he still believed the vitamin reached the body only through food, and that light just activated existing stores. Today we know that the skin itself makes vitamin D in sunlight. MedlinePlus explains that vitamin D "is absorbed from food or produced by the skin when exposed to sunlight", and that rickets is most likely during periods of rapid growth. Mellanby had seen that too: his fastest-growing puppies developed rickets most easily.
Revising this topic for an exam? Our guide to vitamins, their classification and deficiency diseases covers the full list. For another vitamin found by feeding experiments, read about Christiaan Eijkman and beriberi. And if percentage change versus absolute change still trips you up, the Fermi Problem Library has a step-by-step solver that explains the reasoning rather than printing an answer.
What can a student learn from how Mellanby worked?
He started with a diet that reliably produced the disease, so every later result had a clear baseline. He changed one thing at a time, measured bones chemically rather than by eye, and by 1919 had used 200 puppies rather than trusting a handful. And when the evidence on sunlight grew, he tested it himself and changed his view. Being partly wrong did not undo the part he got right.
Try it now: using the numbers above, work out how many grams of calcium oxide puppy 714 still lacked compared with the cod-liver oil puppy, and what percentage of the healthy value it had reached.
