For some, there is confusion between the terms nutritional ketosis that results from following a ketogenic diet and diabetic ketoacidosis (DKA) that sometimes occurs in persons with diabetes, more so in type 1 diabetes (T1D) than in type 2 diabetes (T2D). The purpose of this blog post is to make the distinction between these conditions and to reassure those with diabetes who choose to manage their blood glucose with appropriate insulin doses and a well-formulated ketogenic diet that this combination will not result in DKA. The characteristics of a well-formulated ketogenic diet was described in blog post #13 and my ketogenic diet was described in blog post #9.
In a person with T1D, a low insulin level results from autoimmune destruction of the pancreatic beta-cells that are the sole source of insulin production. If you understand the major functions of insulin, you can understand what goes wrong in persons with T1D. Insulin is an anabolic hormone that directs both the utilization and storage of nutrients (glucose, amino acids, and fatty acids). For example, insulin stimulates the synthesis of glycogen from glucose. Glycogen is a large branching chain of glucose molecules that allows glucose to be stored for later use primarily in muscle and liver. Insulin stimulates the synthesis of triglycerides from both glucose and fatty acids for storage primarily in fat cells (adipocytes). Insulin also stimulates the synthesis of proteins from amino acids to build most of the enzymes, structural, transport, and contractile protein molecules (e.g. muscle) needed for body functions. In addition to promoting the storage of glucose and fat, insulin also instructs cells to use glucose as a fuel preferentially over fat to help clear it from the bloodstream.
In persons with T1D, insulin production declines due to the autoimmune destruction of the pancreatic beta-cells. Lack of insulin results in the diminution of the functions of insulin described above. The body is less able to use glucose resulting in elevated blood glucose (hyperglycemia). In addition, the reduction in insulin is a signal to the liver and kidneys make glucose (gluconeogenesis) which further worsens hyperglycemia. The decrement in insulin is also a signal to the adipocytes to release fat in the form of fatty acids into the blood. The fatty acids travel to the liver which under the condition of low insulin levels makes ketones from the fatty acids. Thus, the liver is making both glucose (gluconeogenesis) and ketones (ketogenesis). These processes are described in more detail in my blog post #6. Protein synthesis is also impaired resulting in muscle wasting which explains why persons with T1D can become emaciated despite increased appetite and food consumption. If insulin production and insulin levels reach a critically low level, the liver gets a strong signal to produce large amounts of both ketones and glucose. The result is a very high level of blood ketones (hyperketonemia) and glucose (hyperglycemia). Because ketones are organic acids, hyperketonemia can result in the dangerous condition, DKA, which also results in (except in rare cases) the presence of hyperglycemia, metabolic acidosis, and dehydration. In addition to acidosis, the other dangerous aspect of DKA is severe dehydration that results from hyperglycemia and hyperketonemia and the kidneys’ inability to reabsorb filtered glucose and ketones resulting in large amounts of glucose (glycosuria) and ketones (ketonuria) in the urine. The glucose and ketones in the urine are osmotic agents that force water to be excreted in the urine resulting in dehydration. It is both the dehydration resulting in dangerously low blood pressure and the acidosis causing changes to how key enzymes and biochemical reactions operate that results in death if not treated in a timely manner. Blood beta-hydroxybutyrate (BHB) ketone levels during DKA are typically in the range of 15 – 25 mM whereas during nutritional ketosis BHB is in the 0.5 – 3 mM range. The low level of BHB during nutritional ketosis does not result in acidosis. Blood glucose levels during DKA exceed 250 mg/dl and often are above 500 mg/dl, whereas during nutritional ketosis blood glucose is near normal due to treatment with insulin injections thus dehydration does not occur. See Figure 1 from my book “The Ketogenic Diet for Type 1 Diabetes” for some of the diagnostic criteria for DKA.
More detailed information about DKA is available in this review article.
The insulin injections also prevent BHB levels from becoming too elevated since insulin inhibits the rate-limiting step in ketone synthesis. As mentioned in blog post #6, I made an effort to see how high I could elevate my blood BHB level by supplementing with coconut oil containing 15% medium-chain triglycerides (MCT). MCT are readily converted to ketones by the liver even in persons not following a ketogenic diet. The maximum BHB level I achieved was 6.9 mM, far below the range seen in DKA. My average BHB level was 2.7 mM, which is in the range seen in nutritional ketosis.
I hope it is clear that DKA is primarily due to insulin deficiency from either undiagnosed T1D or due to stopping insulin injections which occurs in persons with T1D who lack the finances, social support, and/or motivation to make sure they have a continuous supply of insulin. An additional contributing factor in the development of DKA is stress hormone excess due to illness in a person with either T1D or T2D. These illnesses are typically serious, like pneumonia, urinary tract infection, heart attack, etc. which causes release of stress hormones like glucagon, epinephrine, cortisol, and growth hormone resulting in additional production of glucose by the liver and a relative insulin deficient state.
So just to state it clearly, following a well-formulated ketogenic diet and the resulting nutritional ketosis is neither a cause of nor increases the likelihood of developing DKA.
I have found a few case reports in the medical literature by searching PubMed of non-diabetics developing ketoacidosis by following a low carbohydrate diet. In each case, the details of their diet was not given (i.e. likely not a well-formulated ketogenic diet) and they had several days of illness with vomiting and/or not eating prior to presenting to the hospital in ketoacidosis. This means that starvation ketosis/ketoacidosis and stress hormones were likely the causes of or significantly contributing to their problem, not just the diet. I have not found any case reports of diabetics developing DKA due to a ketogenic diet. Simply put, DKA does not occur as a result of following a well-formulated ketogenic diet. I personally know at least ten other doctors who have been following a well-formulated ketogenic diet for 10 – 20 years without any complications. Richard K. Bernstein, MD whom I discussed in blog post #13 has followed a ketogenic diet for his T1D for 45 years without any problems. It would be reassuring to have the results of a large long-term clinical trial using a well-formulated ketogenic diet in the treatment of T1D to know if there are any long-term adverse risks. Such a study would have to be funded by a government or wealthy philanthropist(s) since pharmaceutical companies would have no interest in a dietary therapy. It will likely be many years for such a trial to be conducted if ever. In the meantime, I think the immediate benefits of a well-formulated ketogenic diet in the treatment of T1D is worth taking the risk of an unknown adverse event many years in the future.