A Hundred Years of Treating Diabetes with a Low Carb and Ketogenic Diet

A hundred years ago, Dr. Russell M. Wilder and two Dietitians from the Mayo Clinic wrote a 69-page book titled “A Primer for Diabetic Patients – A Brief Outline of The Principles of Diabetic Treatment, Sample Menus and Food Tables[1]” which outlined the treatment of diabetes using different levels of a low carbohydrate and very low carbohydrate (ketogenic) diet, as well as short periods (12-48 hours) of fasting.

“The effect of ketonemia on the course of epilepsy” (1921), Russell M. Wilder

So how is it that until yesterday, I only knew of Dr. Russell M. Wilder as the physician who in 1921 proposed that a very low carbohydrate high fat diet which resulted in the body’s production of ketones could be used in the treatment of epilepsy in order to produce the same benefits as fasting[2]?

This is the first of two articles on the origins of a ketogenic diet for the treatment of diabetes. The second one can be read here

It was Wilder himself who is credited with coining the term “ketogenic diet” — and his version of the classic 4:1 ketogenic diet (KD) is still used today in the management of epilepsy, as well as in adjunct treatment for glioblastoma (a very aggressive type of brain cancer) — along with chemo and radiation, as well as in some neurological disorders.  But his first work using the diet was in diabetes.

How is it that Wilder is so famous for being a very early implementer of a ketogenic diet for epilepsy, yet virtually unknown for his earlier use of low carbohydrate and ketogenic diet for the treatment of diabetes? Could it be that the discovery of insulin a little later that same year by Dr. Fredrick Banting and medical student Charles Best[3], and its manufacture by Eli Lilly [3] relegated the use of a low carbohydrate diet in diabetes to the pages of history?

Note (February 16, 2021): As I outline below, that is exactly what happened.

While it makes total sense that use of insulin as treatment of those with type 1 diabetes (where a person’s pancreas produces little or no insulin) took a front and center role to keep them from literally wasting away without it, it is unfathomable to me that Wilder’s dietary recommendations did not continue to be widely used in the management of type 2 diabetes —  an impairment in the way the body uses glucose and which results in too much sugar circulating in the blood. Perhaps it was because type 2 diabetes wasn’t identified as being entirely different than type 1 diabetes until Harold Percival Himsworth differentiated between the two in 1936.

Oral diabetes medications such as Metformin and other biguanide derivatives, as well as sulfonylurea such as Carbutamide and Tolbutamide were not developed until 1955-1956,  so prior to 1955 insulin was the only drug treatment for type 2 diabetes[4]. 

ADDENDUM ( February 16, 2021) – A paper published in 1958 by Dr. Russell Wilder provides some much needed understanding as how the discovery of insulin and its free provision to those with diabetes by the drug company Eli Lilly shaped the use of low carb diets in diabetes treatment.

“Insulin at that time cost five cents a unit in the market. However, the patients in our early cases received theirs gratis (i.e. free) for a period of several years, thanks to the Eli Lilly Company.” (p.247-8 [4])

In Wilder’s 1958 paper, he outlines how the A Primer for Diabetic Patients [1] (the book on which this article is based) went from being a diet-first approach to the treatment of diabetes prior to the discovery of insulin to a diet that was indistinguishable from the carbohydrate and protein rich diet of non-diabetics, as the result of the use of insulin.

“The nine editions of the little book A Primer for Diabetic Patients, the first written in 1921, the last in 1950, provide a panorama of diabetic therapy in that interval. The first printing was based on mimeographed instruction sheets prepared in 1920 for the diabetic patients. We were then following the generally accepted treatment of that time, which was based on the research of Dr. Frederick M. Allen at the Rockefeller Institute in New York. It involved an initial period of starvation and the effort afterward to maintain control of glycosuria by a very rigidly restricted diet and periodic fast days. The second edition (1923 ) introduced insulin and diets made more liberal in fat. The pre-insulin diets were continued because of the cost of insulin—5 cents a unit then—and because of a disinclination to give more of this new drug than was absolutely necessary, since it was not yet known whether ill effects would result from the continued use of insulin. As the years went by, greater and greater liberality was permitted, until, in the later editions of the book, the diets recommended, although still controlled as to composition, provided almost as much protein and carbohydrate as would be contained in the well-selected diets of normal persons [4].”

Perhaps the reason diabetes has been considered a “chronic and progressive disease” is because dietary treatment had been all but forgotten after the discovery of insulin.

In 1921, Wilder understood that;

“Diabetes is a disease which in manifested by excretion of sugar in the urine. This sugar comes from the foods which the patient eats, but which is body, owing to the disease, is unable to utilize.”

and his treatment recommendations (pg.12) were;

suiting the diet to the condition of the patient and feeding no more sugar-forming foods than the patient’s body is able to use.

The concept of “eat what you want and cover it with insulin” simply wasn’t an option for those with diabetes, as insulin hadn’t been discovered yet. Diet was the only choice for managing symptoms of the disease — which begs the question, for those with type 2 diabetes who want to get off of diabetes medications now, why isn’t carbohydrate restriction offered as a choice?

No one denies the safety and efficacy of a ketogenic diet for the treatment of epilepsy, yet many deny that a low carb or ketogenic diet appropriate for those with diabetes — when both have been used safely for 100 years! While there is an absolute need to manage the dosage of oral hypoglycemic medications before and during reduction in the amount of carbohydrate in the diet there is no reason that we cannot support a diet-first, not drug-first approach to diabetes treatment and management when people want.

Determining Carbohydrate Tolerance

Diabetes is at its very essence “carbohydrate intolerance” and Wilder describes ‘tolerance‘ as ‘the amount of sugar-forming foods which a person can eat in twenty-four hours without causing sugar in the urine’.

“The tolerance of a given patient is ascertained by feeding foods of known composition in weighed and gradually increasing amounts.”

“The actual procedure will vary with different patients, but, in general, foods of known composition in weighted amounts are fed, the total intake of carbohydrate, protein and fat being increased very gradually as high as possible without the return of sugar in the urine.”

“Some patients will be found to have low tolerance, others may stand 100 gm. of carbohydrate. Every patient should be treated as an individual case, but for convenience in prescribing diets, the following arbitrary grouping is made:

Group A — tolerance below 40 gm. carbohydrate

Group B — tolerance between 40 and 60 gm. carbohydrate

Group Ctolerance between 60 and 100 gm. carbohydrate

Group Dtolerance above 100 gm. of carbohydrate

Use of Fasting and Protein Sparing Modified Fasts

Wilder’s approach incorporates short fasts of 12 – 24 hours as part of the management of blood sugar and up to 2-days if spilling sugar in the urine (i.e. exceeding their carbohydrate tolerance).

Those in Group A who have carbohydrate tolerance of less than 40 grams of carbohydrate (ketogenic level for women and men) are instructed to “interrupt their diet by a “fast day” once a week” — but its not a complete fast.  They are told to “take liberally of liquids”, including beef broth and coffee or tea.

Those in Group B who have carbohydrate tolerance between 40 and 60 grams of carbohydrate are instructed to “institute weekly days of half-fasts” on which they restrict their diet to 20 grams of carbohydrate, as well as 12 grams of protein and 12 grams of fat.

On the appearance of sugar in the urine at any time, the patient, irrespective of his group, must institute a fast day. If the sugar persists, a second fast day should follow the first.

Of interest,  if sugar persists in the urine Wilder’s recommendations are to institute what would be known today as a “protein sparing modified fast“;

If sugar persists, the patient should return to one-half of his diet, continue on this for a week, and then again try the effect of a fast day. After the urine is again sugar-free, he can return gradually to his previous diet.”

Wilder cautions that “longer fasts should never be attempted outside of an institution”, but it’s important to keep in mind that there was no distinction at this point between type 1 and type 2 diabetes.

There is no fasting protocol for those whose carbohydrate tolerance is between 60 and 100 grams.

Daily Macros

Those in Group A who have a carbohydrate tolerance of less than 40 grams of carbohydrate (ketogenic level for women and men) are instructed to eat breakfast, lunch and dinner such that the value of three such meals has 20 grams of carbohydrate, 70 grams of protein and 100 grams of fat (pg. 30 [1]).

Those in Group B who have carbohydrate tolerance between 40 and 60 grams of carbohydrate are instructed to eat breakfast, lunch and dinner such that the value of three such meals has 40 grams of carbohydrate, 70 grams of protein and 100 grams of fat (pg. 33 [1]).

Those in Group C who have carbohydrate tolerance between 60 and 100 grams of carbohydrate are instructed to eat breakfast, lunch and dinner such that the value of three such meals has 60 grams of carbohydrate, 70 grams of protein and 100 grams of fat (pg. 37 [1]).

Those in Group D who have carbohydrate tolerance above 100 grams of carbohydrate are instructed to eat breakfast, lunch and dinner such that the value of three such meals 100 grams of carbohydrate, 70 grams of protein and 140 grams of fat (pg. 41 [1]).

At 70 grams of protein per day irrespective of a person’s weight or gender, these low carb / ketogenic diets provided plenty of satiety and this amount is above the current DRIs of 46 g protein for the average sedentary woman, and 56g protein per day for the average sedentary man.

Fat sources in the sample menus were butter, cream, cheese and eggs and the fat found in the protein.

Carbohydrate sources in the diets were mainly from recipes for something called “Hepco Cakes” made from eggs, cream, Hepco flour, butter and water or from “Cullu-flour Griddle Cakes” made from eggs, salt, water and cellu-flour. Very low carb meals included low carb vegetables and a little bit of fruit, and the higher carbohydrate meals included low carb vegetables, root vegetables such as onion and beet, as well as a bit of fruit.

Wilder’s Low Carb / Keto Diet for Diabetes – a summary

Wilder understood that diabetes is a disease of carbohydrate intolerance and that each person with diabetes “should be treated as an individual“.

He was aware that some people with diabetes will have very low carbohydrate tolerance of less than 40 grams per day requiring a ketogenic level of intake, while others can tolerate up to 100 grams of carbohydrate per day.

Wilder did not restrict  protein, as he did in the 4:1 ketogenic diet he later developed for epilepsy.

Final Thoughts…

My experience in clinical practice over the last 5 years teaching low carb and ketogenic diets is that each person with type 2 diabetes has different levels of carbohydrate tolerance.

When I first started teaching low carbohydrate and ketogenic diets 5 years ago, unless someone was already on a ketogenic diet, I would start those with type 2 diabetes (who were not on any of the medications previously mentioned) at 130 grams of carbohydrate per day and gradually lower carbohydrates until clinical outcomes were reached.

In the past two or three years I came to the realization that none of my clients with type 2 diabetes were tolerating carbohydrate intakes above 100 grams per day — which interestingly is consistent with Wilder’s categories.


People think of the “keto diet” to treat diabetes as something new, but it has been around for over 100 years. When medications have been around for a long time (such as “ASA” i.e. Aspirin), they are given GRAS status (Generally Recognized As Safe) and considered safe by experts, without the need for additional evaluation.

Given that use of both low carbohydrate and a ketogenic diets in the treatment of diabetes along with short periods of therapeutic fasting was developed over 100 years ago, why is this approach not also generally recognized as safe — with specific qualifiers in place for those taking certain medications such as oral hypoglycemic medications?

For those who insist that a ketogenic diet was first used in the treatment of epilepsy, here is the link to the second part of this article which documents clinical use in diabetes treatment prior to 1916 — and likely what Wilder was referring to in his “… as has long been known” statement in July 1923 when he suggested its use in epilepsy.

More Info?

If you would like more information about how I might be able to support your needs, please have a look under the Services tab or send me a note through the Contact Me form.

To your good health!


Special recognition to Jan Vyjidak of London, England, Founder and CEO at Neslazeno.cz for finding A Primer for Diabetic Patients (1922)!

You can follow me on:

Twitter: https://twitter.com/lchfRD
Facebook: https://www.facebook.com/lchfRD/
Instagram: https://www.instagram.com/lchf_rd

Copyright ©2021 The LCHF Dietitian (a division of BetterByDesign Nutrition Ltd.)

LEGAL NOTICE: The contents of this blog, including text, images and cited statistics as well as all other material contained here (the ”content”) are for information purposes only.  The content is not intended to be a substitute for professional advice, medical diagnosis and/or treatment and is not suitable for self-administration without the knowledge of your physician and regular monitoring by your physician. Do not disregard medical advice and always consult your physician with any questions you may have regarding a medical condition or before implementing anything  you have read or heard in our content.


  1. Wilder RM, Foley MA, Ellithorpe D, A Primer for Diabetic Patients – a brief outline of the principles of diabetic treatment, sample menus and food tables,  The Mayo Clinic, W.B. Saunders Company Publishing, 1922
  2. Wheless JW. History of the ketogenic diet. Epilepsia. 2008 Nov;49 Suppl 8:3-5. doi: 10.1111/j.1528-1167.2008.01821.x. PMID: 19049574.
  3. The History of Insulin, diabetes.co.uk, diabetes.co.uk/insulin/history-of-insulin.html
  4. Krochmal M, 10 Facts About the History of Diabetes, https://type2diabetes.com/living/10-facts-history-diabetes/
  5. Wilder, Russell M. “Recollections and Reflections on Education, Diabetes, Other Metabolic Diseases, and Nutrition in the Mayo Clinic and Associated Hospitals, 1919-50.” Perspectives in Biology and Medicine, vol. 1 no. 3, 1958, p. 237-277. Project MUSEdoi:10.1353/pbm.1958.0019.