Reversing Type 1 Diabetes - Is It Possible? | Chris Kresser
HCTP Banner

Could Type 1 Diabetes Be Reversible After All?

by

Last updated on

Type 1 diabetes is characterized by the loss of insulin-producing 𝛽 cells in the pancreas and has largely been thought to be irreversible—until now. Newly published research suggests that there might be a cure for type 1 diabetes after all. Read on to get all the details.

Is it possible to reverse type 1 diabetes?
A recent study has shown that the drug streptozotocin when combined with a fasting mimicking diet may help reduce type 1 diabetes. iStock.com/RyanKing999

While type 2 diabetes is known to be reversible with diet and lifestyle changes, type 1 diabetes has long been thought to be a permanent condition that requires lifelong insulin dependence. Excitingly, a new study published just last month (1) suggests that a “fasting mimicking diet” could effectively reverse the pathology of type 1 diabetes in mice. While the potential for translating these findings to humans is still unclear, this is such a pivotal study that I wanted to take the time to unpack it piece by piece. First though, a bit of background to set the stage.

What Is a Fasting Mimicking Diet, Anyway?

We know that water-only fasting provides many health benefits, including reduced blood glucose, regeneration of the immune system, and cellular maintenance (2). But prolonged fasting is difficult for most people and can cause adverse effects on physical and mental health due to its extreme nature. Researchers have therefore been attempting to design diets that mimic the physiological benefits of prolonged fasting without the burden of complete food restriction.

Fasting mimicking diet may reverse type 1 diabetes

This type of diet is called a fasting mimicking diet (FMD). It is a very-low-calorie, low-protein, high-fat diet that causes changes in glucose, ketone bodies, and specific growth factors similar to those seen during prolonged water-only fasting. The FMD is characterized by cycles of caloric restriction and refeeding. For example, in mouse models of FMD, researchers restrict the amount of food the mouse has access to for four days, followed by three days of unrestricted feeding every week. In humans, one FMD cycle consists of five days of restriction, and eating resumes as usual for the rest of the month. This is typically repeated for three months (3).

Soon, we’ll jump into the results of the study to look at the intriguing effects of an FMD. But first, let’s briefly review what happens to the body in type 1 diabetes.

Pancreatic Anatomy and Type 1 Diabetes

The pancreas contains regions called islets, which are dense clusters of cells that are responsible for secreting hormones. The two main types of endocrine cells are insulin-producing β cells and glucagon-producing α cells. Pancreatic β cells are among the most sensitive cells to nutrient status.

When you eat a meal, they release insulin, which helps to shuttle glucose from the bloodstream into cells to be used for energy production. Between meals, glucagon helps to maintain a minimum level of glucose in circulation.

Type 1 diabetes is an autoimmune condition in which the body’s own immune system attacks and destroys the insulin-producing β cells of the pancreas. It is widely accepted that β cells in the adult pancreas replicate at an extremely low rate (4, 5), and that new β cell formation occurs very rarely (6). The β cell depletion and resulting loss of insulin secretion characteristic of type 1 diabetes is therefore thought to be irreversible.

Exciting developments in stem cell therapy may have potential for treating type 1 diabetes, but this invasive procedure would require complete removal of the dysfunctional pancreas, stem cell transplant, and activation of a complex genetic program to generate a new one. Enter the fasting mimicking diet.

The Results Are In: FMD Reverses Type 1 Diabetes in Mice

I know this is the part you’re all waiting for, so let’s get right to it. The researchers used a mouse model of type 1 diabetes, in which scientists use high doses of the drug streptozotocin (STZ) to cause depletion of β cells (7). Just five days of treatment with STZ was enough to elevate blood glucose, at which point the researchers started half of the mice on their first FMD cycle. The other half of the mice were left to eat unrestricted.

FMD restores insulin-dependent glucose control
In STZ-treated mice eating as much chow as they pleased, blood glucose levels continued to skyrocket. In contrast, mice receiving regular FMD cycles had blood glucose and insulin levels that returned to nearly normal levels by around day 50. Furthermore, glucose tolerance tests at this time point confirmed that FMD cycles in mice improved the ability to clear excess glucose from the blood.

FMD improves the cytokine profile
They decided to measure other substances in the blood as well. Analyzing immune signaling molecules called cytokines can tell us a lot about how the immune system is interfacing with the rest of the body. In this study, they found that mice on regular FMD cycles had reduced cytokines associated with inflammation and β cell damage (TNFα, IL-12) and increased levels of anti-inflammatory cytokines associated with β cell regeneration (IL-2, IL-10).

FMD triggers the regeneration of pancreatic beta cells
All this is great, but what the researchers really wanted to know was if the improvement in blood glucose control and changes in cytokines was mirrored by functional changes in the pancreas. Using cell staining techniques, they found that STZ treatment resulted in a dramatic (85 percent) decrease in the number of insulin-secreting β cells and an increase of non-hormone-producing cells. But with just a few FMD cycles of caloric restriction and refeeding, many of these non-α/β cells redifferentiated into functional insulin-secreting β cells (1)!

Turning Back Time: Reawakening Embryonic Genes

How was the FMD doing this? The researchers were curious too and wondered if epigenetics were responsible. They ground up some of the pancreatic tissue from the mice and used RNA sequencing to determine which genes were being expressed.

They found that the FMD is able to turn back the clock, promoting a gene expression profile in adult mice that is usually only observed during embryonic and fetal development. This is quite an astonishing finding.

I’ve covered epigenetics before—the idea that expression of certain genes can be turned on or off, depending on what stimuli are present. Essentially, the genetic blueprint for building a pancreas is present in every single cell in the body, from the womb through adulthood, as is the blueprint for every organ and structure of your body. But this genetic blueprint only gets turned on at certain times and in certain cells, when the proper signals are present.

For example, in normal mouse development, pancreatic progenitor cells express the proteins Sox17 and Pdx1 around embryonic day 8.5. Some of these pancreatic progenitors are converted into endocrine precursor cells, which then express the protein Ngn3 from embryonic day 11.5 to 18. These Ngn3-expressing precursors ultimately give rise to all of the islet endocrine cells. These proteins are usually not expressed at all once a mouse reaches adulthood.

However, the results of this study suggest that an FMD can induce expression of these embryonic development and β cell reprogramming markers. When the researchers performed the experiments again but intentionally destroyed the Ngn3 cell lineage, they found that FMD-induced β cell regeneration did not occur. This suggests that epigenetic reprogramming is responsible for the improved glucose tolerance and islet regeneration (1).

Beyond Mice: Regeneration in the Human Pancreas

Okay, so FMD reverses type 1 diabetes in mice. But what about humans? Unfortunately, it’s pretty hard to measure the regeneration of a pancreas in living humans, since we can’t collect human tissue like we can mouse tissue. So instead, the researchers performed ex vivo (outside the body) experiments on cultured human pancreatic islets from both healthy people and type 1 diabetics.

Ingeniously, they separately enrolled five human subjects in an FMD lasting five days and took blood samples at baseline and at day five of the FMD. The post-FMD blood serum samples had higher levels of growth factors and ketone bodies and lower levels of glucose, as expected. The researchers then took the cultured pancreatic islets and bathed them in the collected samples. In both healthy islets and type 1 diabetic islets exposed to the FMD-treated serum, there was a trend toward glucose-dependent induction of embryonic genes Sox2 and Ngn3.

They then tried applying commercially available “fasting mimicking” culture mediums that were low in glucose and serum to the cultured islets. When supplied with just this small amount of glucose, insulin secretion was stimulated in both healthy and diabetic islets. There were also major changes in β cell reprogramming markers (1).

Were We Meant to Eat Three Times a Day?

I’ve written before about the mismatch hypothesis—the idea that our genes have not caught up to our modern lifestyle. Our hunter–gatherer ancestors probably had periodic variation of food scarcity and hunting success and likely rarely ate three times a day. The ability of animals to deal with food deprivation is an adaptive response that is conserved across species. In times of scarcity, a mild atrophy of tissues and organs minimizes energy expenditure. Upon refeeding, the body can build these tissues back up to their normal volume (8).

This raises a few interesting questions: is expression of these “embryonic” genes in adulthood really abnormal? Or is it possible that we are meant to have transient expression of these “embryonic” genes periodically throughout our lifetime?  Could our constantly fed state in most of the Western world be the true “abnormal” gene expression pattern? I certainly hope to see more research in this area, especially in humans.

Answers to Your Burning Questions


What’s the takeaway?
In a mouse model of Type 1 diabetes, a fasting mimicking diet causes a short-term reduction in β cell number, which returns to normal levels after refeeding. This occurs through lineage reprogramming and β cell regeneration, effectively restoring insulin production! Similar changes were seen in a cultured human pancreas, but more research is necessary to confirm that β cell regeneration also occurs in living humans.

What about type 2 diabetes? Since I’ve written before about reversing type 2 diabetes with diet, I decided to focus this article on type 1 diabetes. But the research I unpacked in this article was actually only half of the findings from this influential study. The other half of the experiments used a type 2 diabetes mouse model and showed that FMD was also able to rescue mice from late-stage type 2 diabetes. β cell number and insulin secretion were restored after six to eight cycles of FMD and refeeding (1).

Where can I try an FMD? The FMD, as studied in clinical trials, is currently available from Prolon as a specific package of prepared foods and micronutrients. It is intended to be administered under a doctor’s supervision. It’s very likely that a “homemade” version with a similar number of calories and similar composition of carbohydrate, fat, and protein would have the same effects, but this hasn’t yet been studied in a clinical trial. As I’ve written before, fasting is not right for everyone, so please consult with a healthcare practitioner before trying this at home.

Is it really as simple as an FMD? Maybe, but probably not. If in fact this research translates to humans, an FMD may regenerate your pancreas and restore insulin production in the short term, but it’s not going to stop your immune system from destroying it again. You would still need to address the root cause of disease, which is most likely an underlying food intolerance producing antibodies that are cross-reacting with islet cells. I would surmise that an FMD combined with the Paleo autoimmune protocol might just do the trick, but this remains to be tested.

Where can I dive deeper into the scientific research? Head over to the Kresser Institute blog, where I take an even closer look at the latest medical research. Be sure to check out my recent blog post on the benefits of FMDs for cancer, aging, metabolic disease, cognitive health, and more.

We certainly don’t have all the answers yet. Stay tuned for more discussion on my blog about the benefits of fasting mimicking diets and how to apply them in your life. This is certainly a fascinating topic and one that I’ll continue to cover in future blogs and podcasts as we learn more.

Now I’d like to hear from you. Do you or someone you know have type 1 diabetes? Does this research surprise you? Would you want to try an FMD? Let us know in the comments!

60 Comments

Join the conversation

  1. I did a high carb diet with cereal and fruit juices etc for 6 months and ended up with type 1. That was 2 years ago at the age of 27. Im currently water fasting. Im at nearly at 3 days with no food. I fed myself like a pig for a few days to yet as heavy as possible so my body could eat that. Yeah I used insulin for that. I’m also adding electrolytes to my water and keeping my blood sugar stable with occasional insulin shots. My target is 16 days. I am hoping for the regenration talked about in this article. I have done some research on it. I will stop this if my weight gets too low. I started at 150lbs and I’m now 146lbs. I’m 5 8. I was surviving at 120lbs when I wasn’t using insulin so I should be okay for the 16 days. I read that the true healing period is 8-15 days. And after the first 3 or so days the hunger subsides. I’m not that hungry really but I’m definitely not bouncing with energy.

  2. I am a type I diabetic diagnosed at 13 months old and have just turned 31. This article is highly interesting to me because I had two times in college when it appeared I didn’t need my insulin. By then I had been diagnosed for more than 18 years so it wasn’t a honeymoon phase. I have always wondered if some sort of life changes I was experiencing caused this but I never got answers. My doctors didn’t truly believe me so I never even got so much as a conversation about it. When they saw my pump reports and could tell I really wasn’t giving myself insulin, they were stumped. I think they thought I was giving myself shots or something. But, nevertheless, it has always made me feel like there was real hope out there! I’ll be keeping my eyes on this research as it moves forward. Thanks for sharing!

  3. Interested myself. I actually have food allergies to casein, soy and nightshades, so I cannot do the Prolon FMD. Now that I have completely eliminated these and gluten, my RA has greatly improved. So far, I have only fasted 48 hours at one time. This article motivating me to try for a 5 day fast to see if T1D can really be affected in an old timer like me. I would really love to try a homemade version of the Prolon if anybody can point me to a resource.

  4. Please note that there is a proven method to bring the immune system back in line and stopping the autoimmune process. This is the Coimbra protocoll. After beeing on this protocoll for 6 months, you may start the FMD.

    Good luck.

  5. Hi Chris-I was just diagnosed with T1D, I’m 50 years old, still producing insulin, and otherwise healthy. I’ve been toying with fasting and eat no grains, fruits, sugars or starchy veg. My BG levels have been creeping up over the last year and insulin has been prescribed but I’m seeking alternatives and I have consulted with alternative practitioners who feel that I can wait at least 6 months before starting insulin. I’m planning to get a continuous blood sugar monitor and researching that now. I’d love to try your fasting suggestion with supervision. Let me know if you’re interested in working with me. Thanks! Venus

  6. My 7 year old son is Type 1. We have been working with the mind-body-spirit to heal this. We found a correlation between Type 1 and Celiac Disease…that more people with Type 1 have Celiac Disease and vice versa, so he and I shifted to 95% gluten free diet. We also minimize milk, so I agree with the combination of using Paleo with FMD. Now, how to begin a FMD with a 7 year old? I would do whatever diet with him….but I need some guidelines.

    • I actually looked into this and it sounds like fasting mimicking is not recommended for children because fasting messes with their growth function and could stunt their growth permanently.

  7. My daughter is a type 1 diabetic, diagnosed at age 13, about 9 years ago. At the time of her diagnosis her insulin levels were not checked, only her blood glucose. Her c-peptide was drawn though – and it was absolutely normal. We have had every antibody and genetic testing that is available performed and she has never been found to have antibodies of any kind. It always flusters and frustrates me. I feel as soon as she started injecting insulin, her body stopped creating it. Could a misdiagnosis have resulted in medically induced diabetes?
    Since I am only going by her c-peptide levels it is a mystery if her insulin was normal back then and, in fact, she had some other issue causing high blood sugar. She does poorly on insulin and it never seems to work as it “should.” Another point that makes me think it might have been a metabolism issue instead of a supply issue. I’m curious for any perspective. Thank you.

    • CRBPK,
      I like how you think! In fact, my mind works similarly to yours where I am constantly asking questions and having been researching alternative ways to help my son heal. We are currently using a combination of chiropractic treatment, herbs, GF diet, as well as healing modalities of energy work to heal my son of Type 1. If you would like a fellow parent to talk to and bounce ideas off of, please fell free to contact me at [email protected]. I would definitely appreciate being in contact with another parent who is doing research. -Destiny

    • What’s your daughter’s typical fasting level and her range of glucose levels? Was she ever tested for monogenic diabetes.

  8. Type 1 diabetes is certainly an autoimmune disease, and the readiness of the body to attack pancreatic beta cells persists throughout the life of the patient. This would destroy any spontaneous regrowth of beta cells you could generate, unless you used immunosuppressive drugs, which are themselves more toxic than uncontrolled diabetes, so there would be no point. The tendency of type 1 diabetes to run in families clearly shows that the autoimmune attack on the pancreatic beta cells is genetically conditioned, though it is inherited as a cluster of genes in such a complex way that even identical twins may not be identical for type 1 diabetes development. This points as well to the role of some environmental factor in the development of the disease in genetically susceptible individuals, though the factor is probably so ubiquitous, such as a virus, that it cannot be avoided.