Achieving normal blood sugar levels in type 1 diabetes

A recent study described a group of type 1 diabetic (T1D) patients that have achieved normal blood sugar levels by using a low carbohydrate diet [1]. No therapeutic interventions to date, with the exception of islet transplants, have achieved normal blood sugar levels on a cohort-wide basis. Here, I summarize the data from this low carbohydrate study, highlight the significance of the results, and include a note of caution around implementing this type of diet.

Background

Link between carbohydrates and blood sugar level

Diabetes is a disease characterized by high blood sugar levels. Either the body doesn’t produce enough insulin and/or it doesn’t respond well enough to insulin. Blood sugar levels go up after eating a meal (post-prandial blood sugar), and also during fasting, when the body makes its own glucose to supply other organs (gluconeogenesis). Insulin is released in response to high blood sugar levels, and when it’s missing and/or the body doesn’t respond to it, blood sugar levels stay high.

Carbohydrate content of a meal is a strong predictor of blood sugar level after the meal [2], and post-meal blood sugar levels are thought to contribute most to HbA1c [3]. It therefore seems logical to restrict carbohydrate intake as an approach to achieving lower blood sugar levels. Low carbohydrate diets were used to manage diabetes before the advent of insulin [4]. Insulin made it possible for diabetics to survive while consuming a broad range of foods.

Insulin is not enough

Insulin therapy on its own, however, has not been sufficient to achieve normal blood sugar levels. The average HbA1c of a large T1D cohort has been reported to be 8.4 [5], and only a small fraction of patients (less than 1%) have a normal blood sugar level [5]. Higher blood sugar levels are associated with an increased risk of mortality [6], which makes this problematic.

A functional cure for T1D would mean having normal blood sugar levels. The standard of care for T1D is to check blood sugar throughout the day using finger sticks and inject an appropriate amount of insulin, which depends on many factors including meal content, blood sugar, and time of day. For this reason, it’s difficult to predict and match insulin doses to what is needed to achieve normal blood sugar.

There have been several advances to improve glycemic control, including i) use of a closed loop system or “artificial pancreas,” where a continuous glucose monitor operates in connection with an insulin pump [7], and ii) the “bionic pancreas,” where an additional pancreatic hormone (glucagon) is given to provide more precise blood sugar control [8]. Although both of these approaches have been successful in reducing blood sugar levels, they haven’t on their own been sufficient to achieve non-diabetic blood sugar levels (Figure 1).

Islet transplants represent the exception where it is possible to achieve normal blood sugar levels [9], but issues with transplantation that include long-term use of immune suppressants, loss of graft function, and a shortage of donors, make it impractical to implement this on a larger scale.

So while there have been advances in pancreatic hormone delivery to match blood sugar levels, normal blood sugar levels in T1D remain rare and seemingly elusive.

Study findings

Enter Dr. Richard Bernstein, a type 1 diabetic patient diagnosed at age 12 and who purchased a glucose monitor at a time when their use was rare. He used these measurements to determine which foods and corresponding insulin doses would keep his blood sugar most even. He found that eating low-glycemic foods was key to keeping his blood sugar levels normal [10]. He is currently 84 years old and in excellent health.

The low carbohydrate study characterized an online community [11] that follows the teachings of Dr. Bernstein. 316 T1D patients from this community were studied (131 children, 185 adults). The average HbA1c of this cohort is 5.67, which is in the non-diabetic range (Figure 1).

This study helped address concerns around the safety of a low carbohydrate diet. One concern was low blood sugar, however, patients on this protocol reported lower rates of hypoglycemia compared to their pre-diet status. Patients on the diet also reported fewer diabetic ketoacidosis events and fewer diabetes-related hospitalizations in general.

Another concern was malnourishment and growth stunting in the pediatric cohort. However, the height of this group is slightly above average (0.26 standard deviations above the mean, using World Health Organization standard), and although the authors note a decrease in this number from time of diagnosis, they note that it is consistent with the height deceleration seen in T1D.

Overall, this group has accomplished what is considered rare — normal blood sugar levels in type 1 diabetes. Concerns from the medical community around hypoglycemia and growth stunting were not substantiated by this study. The findings of this study highlight the value of using logic, evidence, and careful self-experimentation, as done by Dr. Bernstein and by these patients in this community, to achieve outcomes that are not expected in the medical community.

Discussion

Although this study is characterized as patients following a low carbohydrate diet, restricting carbohydrates is not the entirety of the protocol recommended by Dr. Bernstein. His protocol puts forth appropriate food substitutions to make up for the change in nutrients (higher protein and fat, with slow-acting carbohydrates from cooked vegetables). Additionally, he recommends specific insulin practices that include splitting of long-acting insulin doses to achieve truer basal levels and the use of regular insulin to cover peaks associated with slower-acting carbohydrates. Finally, he recommends consistency in meal content and timing to allow better prediction of insulin doses, and for family members to follow the same dietary practice.

There is a study that cites six cases where malnourishment occurred in children that were put onto a low carbohydrate diet [12], however, it’s not clear whether appropriate food choices were made to make up for the change in nutrients, whether insulin doses were properly adjusted, and whether family members went on the same protocol, which may have provided protection from starvation. I’ve seen a family member make a sudden dietary change without appropriate food substitutions or changing insulin dosing, and the result was several bouts of hypoglycemia. For this reason, it seems that the most critical aspect of expanding use of this protocol will be in the details of its implementation.

Next steps

1. Understanding best ways to implement this diet.

Evaluate patient stories for implementing a low carbohydrate diet. For example, how eating patterns, diet composition, and insulin doses were changed. A comparison with patients that have not had success on the protocol may highlight the approaches that are most helpful.

2. Testing this in a larger cohort.

Once best practices for implementation have been defined, testing this in a larger cohort will determine how broadly this can be used (for example, whether it works in the most extreme cases of diabetes, such as brittle diabetes).

3. Determining the extent of carbohydrate restriction needed to attain normal HbA1c.

It may be that advances in insulin and pancreatic hormone delivery make it possible to attain normal HbA1c through a less restrictive diet. Other strategies to improve insulin sensitivity, such as time-restricted eating, may also help broaden the range of foods that can be consumed while maintaining normal HbA1c.

4. For it to be acknowledged that normal blood sugar levels are possible in T1D.

Although the American Diabetes Association encourages the consumption of low glycemic foods, it hasn’t stated that normal blood sugar levels are possible in the context of diabetes. Acknowledging that normal blood sugar levels are attainable in diabetes could help motivate patients.

Additional notes

Interestingly, although 87% of patients were satisfied with their own diabetes management, only 50% were satisfied with their professional diabetes care team, and 27% did not discuss their use of the protocol with their care team, citing concerns of disinterest, conflict, and being accused of child abuse.

Also, this cohort seems to be an exceptionally motivated group of individuals. Their pre-diet HbA1c was an average of 7.2, which is lower than the overall T1D average of 8.4. Also, a large portion (45%) of this cohort uses a continuous glucose monitor, compared to 10% in the overall T1D population [5].

References

1. Lennerz et al, 2018. Management of type 1 diabetes with a very low-carbohydrate diet. Pediatrics. [Pubmed] [Full text]

2. Zeevi et al, 2015. Personalized Nutrition by Prediction of Glycemic Responses. Cell. [Pubmed] [Full text]

3. Rohlfing et al, 2002. Defining the relationship between plasma glucose and HbA(1c): analysis of glucose profiles and HbA(1c) in the Diabetes Control and Complications Trial. Diabetes Care. [Pubmed] [Full text]

4. Newburgh and Marsh, 1920. The use of a high fat diet in the treatment of diabetes mellitus. Arch Intern Med. [Full text]

5. T1D Exchange Discover Tool. Accessed prior to November 2018 (dataset no longer publicly accessible). [Link]

6. Lind et al, 2014. Glycemic control and excess mortality in type 1 diabetes. NEJM. [Pubmed] [Full text]

7. Thabit et al, 2015. Home use of an artificial beta cell in type 1 diabetes. NEJM. [Pubmed] [Full text]

8. El-Khatib et al, 2017. Home use of a bihormonal bionic pancreas versus insulin pump therapy in adults with type 1 diabetes: a multicentre randomised crossover trial. Lancet. [Pubmed] [Full text]

9. Shapiro et al, 2000. Islet transplantation in seven patients with type 1 diabetes mellitus using a glucocorticoid-free immunosuppressive regimen. NEJM. [Pubmed] [Full text]

10. Dr. Bernstein’s Diabetes Solution. 2011. 4th edition. [Link]

11. Type One Grit community on Facebook [Link]

12. de Bock et al, 2018. Endocrine and metabolic consequences due to restrictive carbohydrate diets in children with type 1 diabetes: An illustrative case series. Pediatric Diabetes. [Pubmed] [Full text]