In the history of diabetes Frederick Madison Allen (1876-1964) will always be associated with the “starvation diet” which kept some patients with juvenile-onset diabetes alive until they could receive insulin. Fasting was not a new treatment but the novel aspect of Allen’s was his insistence that the severest diabetics should be kept permanently underweight. The previous philosophy had been that, after glycosuria had been abolished by fasting, the patient should be fattened up. Allen received tremendous publicity and support from the foremost diabetes specialist of the time, Elliot Joslin. Yet, it is difficult to work out in retrospect how successful his treatment was because of his habit of attributing deaths to infractions by patients. Critics thought his regimen was heartless and cruel and his biographer Alfred Henderson wrote that he ‘attempted to exercise control over his patients like his laboratory animals’.
Allen was born in Iowa and trained in medicine in California where he served his internship in 1907-8. He also attended the University of Chicago medical school where he was greatly influenced by the physiologists Anton Carlson and George N Stewart. Between 1909 and 1912, he worked in the newly developed Department of Public Health and Hygiene at Harvard, first as a volunteer and then with a poorly paid fellowship which turned into three years intensive research on diabetes. Because his papers were so long, no journal would publish them and he borrowed $5000 from his father to print Studies Concerning Glycosuria and Diabetes (1913), a tome of 1,179 pages in which he gave an exhaustive review of the literature on metabolism in general and diabetes in particular.
One of his early experiments used cats to find out if sustained hyperglycaemia caused diabetic complications. The animals were given daily injections of 100ml dextrose subcutaneously for 17 months. Conclusions from the autopsy were that long-standing hyperglycaemia did not cause lowered resistance to infection, skin troubles, cataract, diabetic symptoms, acidosis or proteinuria. Allen’s work which led to the principle of alimentary rest was based on animal experiments in which he removed varying amounts of the pancreas to produce the equivalents of mild or severe human diabetes. Dogs left with 20 per cent of their pancreas or more did not develop diabetes. The fate of those with 80–90 per cent of their pancreas removed depended on what they ate. On a low-carbohydrate diet, they remained relatively well, like middle-aged humans with diabetes — since Eskimos lived on very little carbohydrate, Allen called this an Eskimo diet. Large amounts of carbohydrate (a Hindu diet) wore out the pancreas and what had originally been mild diabetes turned into the severe pancreatic form. However, if the same animals were fed a high fat diet, the glycosuria disappeared or was greatly reduced. From this Allen decreed that patients should order their lives ‘according to the size of their pancreas’ which basically meant reducing the amount of food until glycosuria disappeared.
In 1914 he was given a junior position at the Rockefeller Institute where there was no shortage of patients since physicians were only too willing to send him their ‘hopeless’ diabetics. His first findings on 44 patients were published in 1915 in the American Journal of the Medical Sciences as ‘Prolonged Fasting in Diabetes’.  Between February 1914 and July 1917, Allen treated 96 patients for an average of 69 days per admission - the longest was 304 days. It was claimed that the series was unselected and ‘ranged from the ignorant shiftless poor to the pampered willful rich’. The main reason for being in hospital was education but far from being confined to bed, it was ‘not unusual for patients to lessen the tedium of treatment by going to concerts etc’. In a meeting in New York in 1916 Allen claimed that ‘many of our patients run up eight flights of stairs at the hospital of our institute twenty times a day. Then they walk eight or ten miles in the open air … thin as they are, we are making athletes of them’.
Opinions about Allen’s treatment were sharply divided. Joslin was extremely enthusiastic and, in 1915 at a meeting of the Association of American Physicians, said, ‘Thanks to Frederick Allen, we no longer nurse diabetics we treat them’. Later the same year he wrote eulogistically that, ‘ The advance in the actual treatment of diabetes mellitus during the twelve months just passed has been greater than in any year since Rollo’s time.…The standards of success of the treatment are so simple that they are within reach of the patient. At one stroke the patient is delivered from medicines, patent and otherwise, sham kinds of treatment, gluten breads and in 99 cases out of 100 of alkalis.’ Others felt that the treatment was cruel and pointless since the patients died anyway. It was also noted that some patients died of starvation rather than diabetes.
In 1919, Allen wrote Total Dietary Regulation in the Treatment of Diabetes, a volume of 646 pages plus charts, which, after considerable editing appeared under the names of Allen, Stillman and Fitz, although Allen claimed that he wrote every word.  It described the course of 76 patients in great detail with voluminous charts. It was reviewed in the British Medical Journal in 1921 where it was said, ‘Allen’s case records have real practical value. The reader can match them with cases from his own experience, and he can see exactly what was done and note the result.’
In 1920, after one of many clashes with his superiors, the Rockefeller Institute sacked Allen. Initially he set up in private practice in New York but what he really wanted was somewhere to combine his clinical practice and research. He eventually found an abandoned estate in Morristown, New Jersey. This had originally cost over $1 million but in its dilapidated state only cost $250,000. Allen somehow managed to get enough money to pay the monthly mortgage of $25,000 and in June 1920 opened his Physiatric Institute with what Henderson describes as ‘ceremonies disproportionate to its tenuous financial foundations.’ It was intended to be a prestigious centre for the treatment of people suffering from diabetes, high blood pressure and Bright’s disease [renal failure]. Rates ranged from $50 to $250 per week and diets were so individualised that each patient was served a unique tray – there were almost as many dietitians as nurses! Allen also solved the problem of publishing his own work by founding the Journal of Metabolic Research which he edited and which ran from 1922 to 1926.
In 1927 he introduced an oral remedy for the treatment of diabetes, ‘Myrtillin’, made from blueberry leaves. It didn’t work and left Allen deeply in debt to Squibb, who sued him in 1930 to recover the money they had lent him. Allen had always been a workaholic loner but he was also a poor business manager. In 1936 the institute was bankrupt although, as Henderson writes, ‘only the eviction notice, nailed clearly on the lawn for all to see, convinced him of the inevitable end. He had two weeks to vacate his institute’. For the last part of his life, he drifted from hospital to hospital trying to do research on hypertension and cancer without funds or support. One of his later contributions in 1941 was Refrigeration in leg surgery, a drugless way of providing local anaesthesia for amputation in poor risk diabetic patients. Allen was not a very attractive person, being described as a ‘square-faced stern-looking man who never smiled’. Henderson writes that, ‘very few ever knew him and very few who thought they understood him really did’. He was taciturn and grew progressively more defensive and argumentative as he got older. He worked seven days a week and his only relaxation was a few hours of tennis on Saturdays when, according to Garfield Duncan, ‘he played with the same precision with which he conducted his research’.
With Joslin, he was for nearly half a century the most fervent advocate of what he described in 1953 as ‘my extreme position in favour of control…before and after the discovery of insulin, with the stated object of preventing both progression and complications.
^ Allen FM. Studies Concerning Glycosuria and Diabetes. Boston, WM Leonard, 1913.
^ Allen FM. Prolonged Fasting in Diabetes. Am J Med Sci 1915;150:480-485.
^ Allen F, Stillman E, Fitz R. Total Dietary Regulation in the Treatment of Diabetes. New York: The Rockefeller Institute for Medical Research; 1919.
^ Duncan GG, Frederick Madison Allen,1879-1964.Trans Assoc Amer Physicians 1966;79:13.