Blood Sugar May Increase After Years on a Ketogenic Diet in Type 2 Diabetes

by

Practitioner’s Preface

Type 2 diabetes has traditionally been viewed as a disorder of insulin secretion and pancreatic beta-cell failure, but can be better understood as a bi-hormonal dysfunction characterized by insufficient insulin released from the pancreatic beta cells to suppress the hypersecretion of glucagon from the pancreatic alpha cells. From a clinical perspective, focusing solely on insulin and dysfunctional beta cells is like trying to fix a broken car by repairing the brakes while ignoring a stuck gas pedal.

The bi-hormonal theory of diabetes is not new. It was first conceptualized by Dr. Roger Unger in 1970 in a publication that provided the evidence that pancreatic alpha-cell dysfunction existed in humans and that the alpha cells of people with diabetes were over-secreting glucagon[1]. That paper was followed in 1975 by Unger and his colleague Lelio Orci publishing the seminal paper on the bi-hormonal hypothesis titled, “Essential Role of Glucagon in the Pathogenesis of Diabetes Mellitus” [2]. The bihormonal theory of diabetes continued to be Unger’s life’s work, which culminated in this 2012 publication [3] that was published only 8 years before his death. In his final paper, Unger stated;

“the hormone glucagon has long been dismissed as a minor contributor to metabolic disease, however, here [in this publication] we propose that glucagon excess, rather than insulin deficiency, is the sine qua non of diabetes.”

According to Unger, glucagon excess is the essential condition, without which diabetes is simply not possible, a concept he outlines in detail in his final publication. Unger’s theory may explain why long-term use of a very-low-carbohydrate / ketogenic diet in type 2 may eventually lead to rising blood sugar levels.

Four Approaches to Type 2 Diabetes Remission 

In 2019, I wrote about the three documented ways to put type 2 diabetes into remission.

  1. a low-calorie energy-deficient diet [6,7,8]
  2. bariatric surgery (especially use of the Roux-en-Y procedure) [9,10]
  3. a ketogenic diet [4,5]

In 2026, we now have a fourth option to put type 2 diabetes into remission, and that is with the use of GLP-1 agonists such as semaglutide, dulaglutide, and tirzepatide [11].

A Well-Designed Ketogenic Diet

Since 2017, and where appropriate, I have supported people seeking remission from the symptoms of type 2 diabetes by following a very-low-carbohydrate/ ketogenic diet. Since this diet is both safe and effective, and in my experience, the hunger that accompanies a low-calorie energy-deficient diet makes it unsustainable long term, a ketogenic diet was and is a viable approach over bariatric surgery, which I believe is best used as a last resort.

Those in my practice who followed a well-designed ketogenic diet had results that mirrored the published studies of Virta Health in their one-year data [4], with an average decrease in HbA1c from 7.6% to 6.3% and an average weight loss of 12%. As in Virta Health’s two-year data [5], those in my practice who continued eating a very-low-carb/ketogenic diet sustained their metabolic improvements at two years.

Even those who followed a low-carbohydrate diet (defined as <130 g carbohydrate/day, but in Meal Plans that I design are usually between 60 and 85 g of carbs per day), significantly improved their metabolic health due to an overall decrease in dietary carbohydrate, along with an avoidance of simple carbs. These dietary changes, combined with lifestyle modifications that supported a reduction in insulin resistance, enabled those more recently diagnosed with type 2 diabetes to achieve full remission of symptoms. Even one or two years later, clients continuing to follow the low-carb Meal Plan that I designed for them continued to experience the improvements in fasting blood glucose, HbA1C, and weight.

Long-Term Ketogenic Diet Data

We have good, solid one- and two-year clinical data in those who have been diagnosed with type 2 diabetes about following a very-low-carbohydrate/ketogenic diet, but at this time, there is very limited long-term data.

1. The five-year Virta Health data showed that 

  • At 1 year: The average HbA1c had dropped significantly from 7.6% to 6.3% [4]

  • At 2 Years: The average HbA1c climbed back up to 6.7% [5].

  • At 5 Years: The average 5-year HbA1c for the completers averaged at ~7.3% [12].

  • 20% of the original participants who continued for five years achieved complete diabetes remission, and 32.5% achieved partial diabetes reversal (HbA1c < 6.5% while taking either no medication or only metformin) [12].

2. In 2023, Dr. David Unwin, a UK-based General Practitioner, published a paper analyzing an 8-year dataset from his real-world NHS practice, Norwood Surgery [13], which documented the following.

  • At the start of collecting cohort data, the average median HbA1c was 7.9% (63 mmol/mol), and across the 8 years, the median HbA1c fell to a sustained median level of 6.4% (46 mmol/mol) [13], and that 51% achieved drug-free remission of type 2 diabetes.

  • The data is presented only as a comparison between the initial baseline HbA1C and the latest follow-up HbA1C of the group at 8 years. No chart or description maps out a standardized year-by-year trajectory across the eight years for the whole cohort.

  • That is, the 8-year timeline (2013 to 2021) represents the total duration that the clinic offered the low-carb program, but patients opted in at completely different times throughout those eight years. Because of the rolling enrollment, the duration of time patients spent on the diet varied wildly, and the paper notes that the average time spent on the diet was 33 months (~2.7 years), though some patients had been tracking for close to the full 8 years.

  • In the publication, Dr. Unwin mentioned that HbA1C tended to drift or worsen as time went on. This insight came from his ongoing clinical observations of individual patient charts over the lifespan of the service evaluation, rather than a year-by-year data table.

  • Dr. Unwin concluded that the increase in HbA!c was the result of “carb creep“ and was tied to a self-reported increase in carbohydrate consumption rather than a failure of the biological mechanism.

When Dietary Adherence to a Ketogenic Diet is Perfect

While “carb creep” may explain glycemic drift for many patients, this data and the five-year Virta Health data do not account for a subset of individuals whose adherence to carbohydrate restriction remained flawless, yet whose blood glucose began to rise anyway. It seems there may be a blind spot in the current literature, one that the bi-hormonal theory of diabetes explains.

Personal Experience Mirroring Study Data

My personal results following a very-low-carb/ketogenic diet from March 2017 to March 2019 mirrored Virta Health’s one and two-year data for fasting blood glucose, HbA1C, and weight loss. I was also able to maintain diabetes remission for several years (March 2019 to March 2020), at which time I was diagnosed with profound hypothyroidism. At that point, my own data was no longer able to be benchmarked to study data. That said, I continued to eat a low-carb diet even during and after thyroid treatment; however, my HbA1c continued to creep up, and most notably, my blood glucose would spike after low-carb, high-protein meals, which baffled me. I wondered if this could be due to my thyroid? Or could it be that this would have happened anyway? I read through the literature, trying to understand a mechanism that would explain it, and I discovered the research of Dr. Roger Unger, which made it understandable. It also led me to the following question;

Could it be that a very-low-carb/ketogenic diet works very well for the first few years in reducing blood glucose and putting the symptoms into remission, but that long-term, there may be insufficient insulin released from the pancreatic beta cells to suppress the hypersecretion of glucagon from the pancreatic alpha cells, which drives the increasing blood glucose?

Not Carb Creep but Alpha Cell Creep in a Long-Term Ketogenic Diet

As clinicians, we expect a very-low-carb diet to reduce insulin needs, but how would low insulin release due to long-term adherence to a very-low-carb/ketogenic diet drive hypersecretion of glucagon from the pancreatic alpha cells?

To understand how this happens, we need to understand what happens at the level of the pancreas.

Blood-Sugar-May-Increase-After-Years-of-a-Ketogenic-Diet-in-Type-2-Diabetes

In people who have never had type 2 diabetes, the pancreatic alpha cells and pancreatic beta cells exist in a tightly regulated paracrine relationship. This is a form of local cellular communication where one cell secretes chemical messengers that diffuse through the extracellular space to affect the nearby target cell. In healthy individuals, when insulin rises, glucagon release is suppressed.

  • In healthy individuals, the beta cells of the pancreas secrete insulin when they eat, which is communicated to the alpha cells of the pancreas and signals them to stop secreting the hormone glucagon. The insulin that is released locally within the pancreas first acts as an immediate, powerful biological “brake” that shuts down glucagon secretion from the alpha cells.

  • In people who have been diagnosed with type 2 diabetes, the beta cells release insufficient amounts of insulin due to beta cell dysfunction. For this reason, a very-low-carbohydrate/ketogenic diet works well for the first few years, because there isn’t the need for much insulin to be released from the beta cells with very low carbohydrate intake.

The problem with gradually rising HbA1C levels and/or rising blood glucose after consuming a low-carb, high-protein meal may arise after following a long-term, very low-carbohydrate diet, due to the reduced daily demand for insulin. This may lead to a state where baseline insulin secretion becomes so low that there is simply not enough local insulin present to suppress the hypersecretion of glucagon from the pancreatic alpha cells. 

Without insulin to quiet the alpha cells, continuous secretion of excess glucagon signals the liver to aggressively ramp up both glycogenolysis (breakdown of glycogen) and gluconeogenesis—the creation of new glucose from non-carbohydrate sources, like amino acids from protein.

This explains why blood glucose can spike after a low-carbohydrate, high-protein meal because the influx of amino acids from the dietary protein strongly stimulates the alpha cells to release glucagon and make new glucose, while at the same time, there is insufficient insulin release to counter it despite there being zero or near-zero carbohydrate intake.

This mechanism suggests that the rising HbA1C that has been observed after years of successful carbohydrate restriction may not always be a sign of patient non-compliance or “carb creep” at all. Instead, it may represent a physiological paradox of the long-term very-low-carbohydrate diet itself: a state of adaptive, relative insulin deficiency that inadvertently unleashes glucagon-driven gluconeogenesis.

Biological Subset of Long-term Non-Responders 

Within Type 2 diabetes studies, there is a distinct biological subset of people who are non-responders over the long term, even when they do everything right, and this comes out of the DiRECT trial protocols and Dr. Roy Taylor’s team [6, 7, 8], where they identified that 10% to 15% of patients are “non-responders” due to their underlying pancreatic biology.

The literature defines this 10% to 15% phenotype by two distinct characteristics:

  1. The “First-Phase Insulin Response” Failure: In the responders, clearing fat out of the pancreas allows the beta cells to recover their rapid, first-phase insulin spike. But in this 10% to 15% phenotype, the beta cells have crossed a “point of no return” of dedifferentiation. They simply could not secrete that rapid burst of local insulin anymore.

  2. The Unbraked Alpha Cell: Because this phenotype lacks that first-phase local insulin response, they completely lack the paracrine “brake” needed to quiet the alpha cells.

When you put this specific 10% to 15% phenotype on a long-term, very-low-carbohydrate/ketogenic diet, their baseline insulin drops even lower. Because their beta cells are biologically incapable of delivering that localized, high-concentration insulin burst to suppress the neighboring alpha cells, their alpha cells release glucagon uninhibited. This is why these individuals don’t sustain remission. It isn’t a lack of willpower, or “carb creep,” but a fixed genetic and structural phenotype where the alpha cells are left entirely unmoderated.

This raises the question as to whether only 10-15% of people on a long-term very-low-carbohydrate / ketogenic diet will experience alpha-cell overactivity due to a lack of a first-phase insulin response, or whether the amount is larger than that, and the 10-15% represents a subset. More study is certainly needed,

Clinical Application

By viewing the adaptive, relative insulin deficiency of type 2 diabetes through Roger Unger’s bi-hormonal lens, it can be seen that, at least in some individuals, maintaining long-term remission of type 2 diabetes symptoms requires more than keeping dietary carbohydrate to 30-50g per day (ketogenic levels).

While we have ample clinical data to support that significant dietary carbohydrate restriction is both safe and effective, and works very well in the first year or two to put the symptoms of type 2 diabetes into remission, the limited long-term clinical data that we have from Virta Health may indicate that, in a subset of individuals, focusing only on reducing insulin is like trying to fix a broken car by repairing the brakes, while not repairing a stuck gas pedal.

For such individuals, the solution is elegantly simple and focuses on dietary changes that promote the release of a very small amount of insulin from the pancreatic beta cells at each meal. This will act on the  “stuck gas pedal” of pancreatic alpha cells, preventing significant endogenous production of glucose, preventing rising HbA1C levels, and blood glucose spikes after consuming a very-low-carbohydrate, high-protein meal.

Such a solution shifts the diet from very low-carb/ketogenic after an initial year or two, to a low-carbohydrate diet with only enough slowly metabolized carbohydrate to simulate a first-phase insulin response, to inhibit the pancreatic alpha cells from driving unregulated gluconeogenesis. I will write about these dietary changes in the days ahead.

More Info

If you would like dietary support to better manage blood sugar levels or to put symptoms of type 2 diabetes into remission, you can learn about me and the Comprehensive Dietary Package that I offer.

To your good health.

Joy

You can follow me on:

Twitter: https://twitter.com/jyerdile
Facebook: https://www.facebook.com/BetterByDesignNutrition/

 

Support for Type 2 Diabetes

I can design a Meal Plan for you to address high circulating levels of insulin and the underlying insulin resistance of type 2 diabetes
which will enable you to lose weight, without being hungry all the time.

Book an Appointment

 

References

  1. Unger RH, Aguilar-Parada E, Müller WA, Eisentraut AM. Studies of pancreatic alpha cell function in normal and diabetic subjects. J Clin Invest. 1970 Apr;49(4):837-48. doi: 10.1172/JCI106297. PMID: 4986215; PMCID: PMC322540. https://doi.org/10.1172/JCI106304
  2. Unger RH, Orci L. The essential role of glucagon in the pathogenesis of diabetes mellitus. Lancet. 1975 Jan 4;1(7897):14-6. doi: 10.1016/s0140-6736(75)92375-2. PMID: 46337.
  3. Unger RH, Cherrington AD. Glucagonocentric restructuring of diabetes: a pathophysiologic and therapeutic makeover. J Clin Invest. 2012 Jan;122(1):4-12. doi: 10.1172/JCI60016. Epub 2012 Jan 3. PMID: 22214853; PMCID: PMC3248306.
  4. Hallberg SJ, McKenzie AL, Williams PT, Bhanpuri NH, Peters AL, Campbell WW, Hazbun TL, Volk BM, McCarter JP, Phinney SD, Volek JS. Effectiveness and Safety of a Novel Care Model for the Management of Type 2 Diabetes at 1 Year: An Open-Label, Non-Randomized, Controlled Study.
  5. Athinarayanan, S. J., Adams, R. N., Hallberg, S. J., McKenzie, A. L., Bhanpuri, N. H., Campbell, W. W., Volek, J. S., Phinney, S. D., & McCarter, J. P. (2019). Long-term effects of a novel continuous remote care intervention including nutritional ketosis for the management of type 2 diabetes: A 2-year non-randomized clinical trial. Frontiers in Endocrinology, 10, 348. https://doi.org/10.3389/fendo.2019.00348
  6. Lim EL, Hollingsworth KG, Aribisala BS, Chen MJ, Mathers JC, Taylor R. Reversal of type 2 diabetes: normalisation of beta cell function in association with decreased pancreas and liver triacylglycerol. Diabetologia2011;54:2506-14. doi:10.1007/s00125-011-2204-7 pmid:21656330
  7. Steven S, Hollingsworth KG, Al-Mrabeh A, et al. Very low-calorie diet and 6 months of weight stability in type 2 diabetes: pathophysiological changes in responders and nonresponders. Diabetes Care2016;39:808-15. doi:10.2337/dc15-1942 pmid:27002059
  8. Lean ME, Leslie WS, Barnes AC, et al. Primary care-led weight management for remission of type 2 diabetes (DiRECT): an open-label, cluster-randomised trial. Lancet2018;391:541-51
  9. Cummings DE, Rubino F (2018) Metabolic surgery for the treatment of type 2 diabetes in obese individuals. Diabetologia 61(2):257—264.
  10. Madsen, L.R., Baggesen, L.M., Richelsen, B. et al. Effect of Roux-en-Y gastric bypass surgery on diabetes remission and complications in individuals with type 2 diabetes: a Danish population-based matched cohort study, Diabetologia (2019) 62: 611. https://doi.org/10.1007/s00125-019-4816-2
  11. Fadini GP, Giaccari A, Broglio F, Nollino L, Fattor B, Anichini R, Meregalli G, Avogaro A, Consoli A; GLIMPLES Study Investigators. Type 2 diabetes remission after initiation of GLP-1 receptor agonists: frequency, characteristics, and outcomes using multiple definitions in an observational study. Lancet Reg Health Eur. 2025 Oct 15;59:101499. doi: 10.1016/j.lanepe.2025.101499. PMID: 41142657; PMCID: PMC12550164.
  12. McKenzie AL, Athinarayanan SJ, Van Tieghem MR, Volk BM, Roberts CGP, Adams RN, Volek JS, Phinney SD, Hallberg SJ. 5-Year effects of a novel continuous remote care model with carbohydrate-restricted nutrition therapy including nutritional ketosis in type 2 diabetes: An extension study. Diabetes Res Clin Pract. 2024 Nov;217:111898. doi: 10.1016/j.diabres.2024.111898. Epub 2024 Oct 20. PMID: 39433217.
  13. Unwin D, Delon C, Unwin J, Tobin S, Taylor R. What predicts drug-free type 2 diabetes remission? Insights from an 8-year general practice service evaluation of a lower carbohydrate diet with weight loss. BMJ Nutrition, Prevention & Health. 2023;6:. https://doi.org/10.1136/bmjnph-2022-000544 Diabetes Ther. 2018 Apr;9(2):583-612. doi: 10.1007/s13300-018-0373-9. Epub 2018 Feb 7.