The Ketogenic Diet and Diabetes: The Virta Health Study. Podcast with Stephan Guyenet

The number of people with diabetes worldwide increased nearly four-fold from 1980 to 2014, and about 9% of people worldwide now suffer the consequences of this debilitating disease. People with diabetes are at increased risk of many ailments, including cardiovascular disease and kidney failure. With this in mind, many people are desperately seeking a cure for diabetes. Recently, interest has surged in the ketogenic diet and diabetes.

In today’s show, I’m pleased to welcome back Dr. Stephan Guyenet to discuss a recent study on the effectiveness and safety of treating type-two diabetes with a ketogenic diet intervention. Long-time listeners may recall my previous conversations with Stephan, in which we talked about why we are fatter than our ancestors and whether high-protein diets are actually bad during weight loss efforts. Pro users of humanOS.me will also know Stephan from the Ideal Weight Program that we collaborated on, and he is also the author of the book, The Hungry Brain – Outsmarting the Instincts That Make Us Overeat. Stephan spent 12 years researching the neuroscience underlying neurodegenerative disease and body fat regulation, and you can find more of his excellent work here.

Just last month, a study led by Dr. Jeff Volek of Ohio State University compared the effects of a one-year ketogenic diet and diabetes support program to the type of care that diabetes patients routinely get from their healthcare providers. Participants had had type-two diabetes for about eight years before the study and were obese – about 220 lbs, on average. Patients in the ketogenic group were not told to consciously reduce their calorie intakes, but they were instructed to eat less than 30 grams of carbohydrate each day. As Stephan notes, “that’s the equivalent of eating an apple and a half a day”.

What did those in the ketogenic diet intervention experience?

Very impressive! And these results were much better than those of participants in the group receiving usual care. But is there more to the story?

There sure is.

First, the ketogenic intervention was the program run by Virta Health, a company that several of the authors work for. As Stephan notes that “There’s a huge conflict of interest in this study because the researchers are testing a program that’s a major source of revenue for them”. This said, he adds that this alone is not enough to disregard the study, we just need to scrutinize it more closely.

Also, the study was not a randomized controlled trial. The people in the ketogenic group chose to be in that group of the study and paid a lot of money to participate. It’s very likely that there was a large difference in motivation and support between the ketogenic diet group and the standard care group. While these differences between groups don’t detract from how impressive the results of the ketogenic diet group were, they do make it harder to determine the precise effects of the ketogenic diet.

If we really wanted to see, for instance, the difference between just the two diets in diabetes, the study would randomize the patients into one of the two groups and try to provide the exact same support to both groups. But this study was essentially designed in a way that would magnify the difference between the standard of care treatment and the ketogenic diet group.

This said, the study was pre-registered, meaning that the scientists published exactly what they planned to do before the study, preventing them from dredging through the data, searching for results to make their protocol look as good as possible.

A few other things are noteworthy. First, when the researchers looked at participants’ ketone levels (to measure of compliance with the intervention), they found that adherence was excellent at first. But compliance had decreased dramatically after a year, indicating that participants may have struggled to stick to the stringent dietary guidelines. Nevertheless, the study unambiguously showed that the ketogenic diet helped patients regulate their blood glucose levels more effectively, so the intervention was very effective, even if dietary adherence became worse over time.

After the show ended, Stephan and I continued to discuss the study and findings for a bit. We both agreed that while adherence slipped, the participants were still producing ketones and seeing results. So, despite the fact that we discussed in the show how this lack of adherence was an indication that this diet might be hard for the average person with diabetes, it is possible that people with diabetes don’t have to comply perfectly with the most stringent version of a ketogenic diet to experience weight loss and improvements in diabetes symptoms.

So, the ketogenic diet seems to be a useful strategy for some people with type-two diabetes. And as Stephan notes, “this certainly is a step… toward demonstrating that ketogenic diets are safe”.

But how have other people portrayed the study in the media?

Can we say that the program “reversed” diabetes?

And is it better than alternative dietary strategies, some of which have recently been shown to be remarkably effective? (Here is a link to a relevant previous post – search for the section entitled: A very low-calorie diet may reverse type 2 diabetes).

For answers to these questions and more, check out the interview using the links below!

On Soundcloud, iTunes, Google Play, Stitcher, and YouTube

Dan prepared for and conducted the interview, Greg wrote the first draft of this blog post, and Dan edited the draft.

Want us to produce more information about the ketogenic diet and diabetes? You can help support the work of the humanOS team by making a donation. If you do, thank you so much – we really appreciate your support! And no worries if you can’t chip in, the important thing is that you benefited from listening today.

Stephan Guyenet - 00:06: This study showed that the ketogenic diet controls blood glucose levels relatively effectively, relative to other strategies and it's one of the more effective diet and lifestyle strategies for doing so.

Kendall Kendrick - 00:24: humanOS. Learn. Master. Achieve.

Dan Pardi - 00:34: Hello, everybody. Welcome back to humanOS Radio. As you well know topics that relate to weight and metabolism are frequently discussed on the show. We've also had shows on ketosis in the past specifically when I spoke to Neil Copes about the beneficial properties of the ketone beta-hydroxybutyrate in relation to better aging and recently there was a study looking at a diabetes management program utilizing the ketogenic diet versus standard of care. The results were very impressive as you'll see but there are some points of concern. One of those concerns revolves around claims that have been made about the findings.

01:06: I've asked frequent guest, Dr. Stephan Guyenet, back to the show to join me to discuss this. Without further ado, Stephan, welcome back. How are you?

Stephan Guyenet - 01:13: Thanks, Dan. It's good to be here. I'm doing well. Thank you.

Dan Pardi - 01:16: Great well to get into this if you could give us a little detail about this recent study that came out in diabetes therapy. Who were the designers of the study and what company are they are part of? Give us some detail there.

Stephan Guyenet - 01:26: Yeah so the study was published in February in the journal, Diabetes Therapy. It was conducted by Stephen Phinney, Jeff Volek and colleagues. These people are longtime researchers and proponents of low carbohydrate diets and they run a commercial diabetes management program called Virta Health.

01:47: I went and checked out their website. It's really slick. The program costs $370 per month and for that you get intensive support from a staff of doctors. You get diet advice and again intensive support for diet adherence. You get equipment like a wireless scale and a glucose meter and ketone tests strips and they work with your local primary care physician.

02:11: The intervention is based around a very low carbohydrate ketogenic diet with intensive support to keep people on it. Just to make sure everyone's up to speed on what a ketogenic diet is, it is a diet that's very low in carbohydrate and high in fat that essentially causes the body to shift from burning carbohydrate predominantly to burning almost exclusively fat. The way that happens normally you have glucose in your blood that comes from the starch you eat and the sugar you eat. You have blood sugar and that blood sugar is used by all your tissues to fuel all of your normal metabolic processes. The brain in particular is heavily dependent on glucose. That's almost its sole metabolic fuel but the brain can also use an alternative fuel source called ketones. What happens when you're not eating a lot of carbohydrate, your liver starts taking the fat that you're eating and the fat from your body fat stores and starts converting it into what's called ketone bodies such as beta-hydroxybutyrate, aceto acetate, and acetone and those in your circulation can be used by all of your tissues including your brain.

03:22: It's a mechanism that the body has for adapting to periods of low carbohydrate intake whether that is because you are fasting or starving or eating a very low carbohydrate diet. To get there, you have to eat roughly somewhere in the neighborhood of 20 to 50 grams of carbohydrate a day. In this particular study, people were eating less than 30 grams of carbohydrate a day or at least that was the target. Just to give you a sense of what that looks like that means that the only carbohydrate, that's the equivalent of eating an apple and a half a day and that's the only carbohydrate you eat for your entire day. Everything else is non-starchy vegetables, no fruit, no bread, no grains, no beans, nothing like that. It's very focused on foods that are higher in fat and animal foods although there are also of course plant foods that are high in fat like avocados and nuts.

04:20: One of the things that I really want to emphasize about this intervention is it didn't involved deliberate calorie restrictions. People were just eating this carbohydrate restricted diet, this ketogenic diet that causes ketone production and however much they ate to fullness on that food was however much they ate. They weren't counting calories. They weren't restricting calories and so the study ... I just described the Virta Health program broadly speaking. The study was a test of the Virta Health program versus usual diabetes care and just to describe what that means usual care just means that people are going along with their typical diabetes care through their normal doctor. It's much less intensive. The doctors periodically checking their blood glucose. Maybe they have a meeting with a dietitian about mainstream diabetes eating plans every now and then. Basically, it was a much less intensive type of intervention and a much more conventional type of intervention that it was being compared to.

05:23: I want to mention that there's a huge conflict of interest in this study because the researchers are testing a program that's a major source of revenue for them. These are the same people studying Virta Health that are running Virta Health and this is a for profit organization. I think we have to recognize that but at the same time in my opinion that alone is not enough to disregard the study. I don't think you can just look at a conflict of interest and say, "Well, I'm just going to ignore this study because there's a conflict of interest."

Dan Pardi - 05:54: Yeah.

Stephan Guyenet - 05:54: I think personally, the most reasonable way to interpret that is to say, "Okay, I'm gonna be more skeptical when I'm looking at this than I would otherwise be but I'm not just gonna completely disregard this."

06:07: 349 people with type 2 diabetes got to choose whether they wanted to be in the Virta Health group, the ketogenic diet group, or the care's usual group and at baseline these people were very heavy. They weighed an average of about 110 kilos or 240 pounds. They had a body mass index of about 37 to 40 particularly in the Virta Health group. They had a BMI of about 40 so the cutoff for obesity is 30. The cutoff for overweight is 25. On average, these people had a lot of excess fat mass and they had diabetes for about eight years. These are people with longterm diabetes but on average it was fairly well controlled at baseline using a good amount of drug intervention. Most of these people were on metformin, sulfonylureas, they were on other antidiabetic drugs but in that they were fairly well controlled on average.

07:05: In the Virta Health group you have people who have diabetes, facing potentially nasty complications, these people chose to be on the ketogenic diet. They were not randomized to the ketogenic diet and they're paying a lot of money to participate in this intervention. These people are going to be highly motivated and highly enthusiastic to adhere to this diet and to participate in this study. This is comparing them to people who are just going along with their usual diabetes care and who chose to do their usual diabetes care.

07:37: I'm just trying to emphasize the importance of the fact that this was not a randomized controlled trial. People were not randomized to both groups and so we can't really say that these two groups were truly comparable to one another. Probably there were many differences between these people including in how motivated they were to do something about their diabetes and that doesn't mean the results are invalid. In particularly like a study like these where the effect sizes are really large, you can't just say, "Oh, well it wasn't randomized so I'm not gonna pay any attention to this." It is still a useful, informative result but it does warrant a greater degree of skepticism because again the two groups we're comparing are not truly comparable and non-randomized studies historically have tended to report larger results than randomized studies. Basically, the lower you go down on the chain of study rigorousness, the more likely you are going to be to get these really big knock your socks off types of effects that might not be very easy to replicate by a true randomized design.

08:47: The primary outcomes of this study were hemoglobin A1c, which is a marker of glucose control and this is a good thing to focus on in a diabetes study, body weight another useful thing to focus on particularly since these people were very heavy and the use of diabetes medication. Those were the three primary outcomes and the diabetes medications seeing how much they're changing basically. Are people able to get off their insulin? Are they able to get off the metformin? Are they able to get off the sulfonylureas, et cetera, et cetera?

09:19: I want to make a really important point here that the study was preregistered. The researchers declared in advance what their research plan was going to be and they declared, which outcomes are going to be the most important in their analysis. This is really, really important. I really want to emphasize this. Listeners may or may not be aware of what's happening right now with the so-called replication crisis in science but basically, we're finding out that a lot of studies that had been done are not as solid as we thought they were and essentially science is more uncertain than progressing at a less efficient pace than we thought it was because of limitations to the methods that people are using.

10:02: One of the most powerful weapons to combat that and to make science more rigorous is to preregister your studies in advance. You say in advance what you're going to look at, what kind of analyses you're going to run and, which outcomes are going to be most important. Basically, what that prevents you from doing is fishing around in your results and just looking around till you find something that's statistically significant and then focusing on that in your study. That strategy yields a very high risk of false positive outcomes that are misleading and cannot be replicated. These people went through the trouble of preregistering the study so that they could say to everyone, "Look we said exactly what we're gonna look at in the study in advance so you can be confident that we didn't do that. We didn't just root around looking for significant things trying to find a desired outcome."

Dan Pardi - 10:54: Right. I know a lot of journals now, they won't even allow you to submit your paper unless the study has been preregistered where you declare what you're looking for in advance. That's a positive change. I'm glad this study did that as well.

11:07: I'll just do a quick summary because that was really rich and useful. In this study, a commercial diabetes management program from Virta Health that does intensive support, uses a ketogenic diet does tracking and measuring was compared to standard of care but it wasn't randomized and that they let the participants choose, which group they wanted to be in and so that automatically can set up between groups because one group is just coming along as they were. The other group is saying, "Okay, I want to be in this group." You could clearly see that there could be a difference then in motivation between the group. While we still are interested to see the magnitude of effect, making comparisons between let's say standard of care and any other intervention that let's people select into it and gives a lot more care, and attention to how the people are performing that needs to be considered. Good points that you made there. What did they find?

Stephan Guyenet - 11:56: We'll start of by focusing on the primary outcomes because those are the most important things. Those are the things that they declared in advance were going to be the most important.

12:05: The hemoglobin A1c declined a lot more in the ketogenic diet group, the Virta Health group, declined by over 1%, which is a large effect compared to other diet and lifestyle modification studies. It's not head and shoulders above everything else that's been seen but it's definitely on the upper end of what we've seen so it was an effective diet and lifestyle program for reducing HbA1c.

Dan Pardi - 12:34: Can you tell us a little bit about HbA1c for somebody's not familiar with it?

Stephan Guyenet - 12:38: Yeah, thanks for reminding me. I won't get into the molecular details of it but basically what it is it's a measure of your average blood glucose level over the past three months or so. It's a marker you can measure in the blood that is proportional to your average blood glucose levels. Since diabetes is diagnosed by having high blood sugar and that's a key aspect of the disease, then hemoglobin A1c is a key measure of the disease. Having high blood glucose is considered to be harmful as well to your organ systems.

13:12: If your HbA1c is going down, that's a good sign that you're not getting as much uncontrolled blood glucose release and that you're probably not going to get the same level of organ damage over time as someone who has less well controlled glucose.

13:28: Another thing I want to point out is that the hemoglobin A1c declined and the decline was impressive but it did not get down into the normal range. On average, these people did not have a level of hemoglobin A1c that would indicate that their blood sugar levels were within the normal healthy range. Even despite the fact that they're eating almost no carbohydrate, they're still experiencing levels of blood sugar above where you would want. This isn't really a critique just to clarify. It definitely improved and these are people who've had long term diabetes. That's not a critique but I just want to point out that this is not a miracle cure that brings you back to exactly where you want to be.

Dan Pardi - 14:09: Yeah, it's important context.

Stephan Guyenet - 14:11: Yeah, and one of the other interesting things is that the hemoglobin A1c declined the most and people who needed it the most. People who've had highest at baseline had the greatest declines over time so that's good. That's what you would want to see.

14:25: Second primary outcome is body weight, another very important one. The ketogenic diet group, the Virta Health group lost 12% of their body weight while the body weight of the control group did not change. I would say that this is probably the most impressive result.

14:42: By the way, I'm not sure I mentioned before, all this outcomes is at one year. After one year on this intervention, that's what I'm talking about. If you look at the curve, essentially they lost weight continuously over most of that year and experienced very little or no rebound on average. Over a one year period a 12% loss of weight is quite impressive compared to other weight loss programs.

15:07: Again, you have to factor in the fact that this is not a randomized trial so the effect is probably bigger than what you would see in a randomized trial. You can't really directly compare to what's been shown in randomized control trials but nevertheless, it does show that people who go into this program lose 12% of body weight over a year. That is impressive. Five percent weight loss is considered clinically significant and can be associated with major metabolic improvement so 12% is very good.

Dan Pardi - 15:35: I think you said the average weight was around 240 coming into the trial so that's like a 30 pound loss, close to that. That's pretty good.

Stephan Guyenet - 15:41: Yeah, something like that. I don't remember exactly off the top of my head but yeah and then the diabetes medication was the third primary outcome. Again this is an important one, you would hope that people would get off their medications or use less medications if a diet and lifestyle intervention is effective. The reporting was a little bit squirly here because they separated metformin from non-metformin drugs and I didn't see that in the preregistration. I think that could be a case where they just diced it up in a way that made it look more favorable, but there was a substantial decline in the use of certain types of drugs.

16:20: For example, insulin use, injected insulin is obviously a key drug for controlling diabetes was cut nearly in half. Half of people were no longer on insulin at the end so that's good. Most people in both groups were still on metformin at the end. There was still plenty of drug use but as far as diabetes drugs go, metformin is really one of the best, cheapest and most effective drug. Getting people off a lot of these other drugs including insulin, insulin's a pain. It's expensive so that's I would say more important than getting people off metformin in my opinion.

16:56: Then there were some other effects that were not primary outcomes but they measured a whole bunch of different things and I'll just cover some of the ones that I think are most interesting. People experience reductions in blood pressure, reductions in an inflammatory marker called C-reactive protein, reductions in fasting insulin level and glucose. These are all things that are expected with weight loss and very low carbohydrate diet but never the less they did observe those. They also reported a reduction in a measure of insulin resistance. This is an indirect measurement called HOMA-IR. The idea would be that their insulin sensitivity improved but in this context it's not really convincing because it's based on fasting insulin, lower carbohydrate diets, lower fasting insulin. It's actually not really an informative measure of insulin sensitivity in this context. We don't really know how their insulin sensitivity changed. However, I would predict that it improved because they lost a bunch of weight and generally weight loss improves insulin sensitivity. If true, that would be a very good thing.

18:02: I would say the one negative thing that you can find in this paper is that the LDL level, LDL or bad cholesterol increased by 10% in the Virta Health group. That's a pretty meaningful increase. However, one thing that I will note is that when they looked at their LDL particle count, in other words not just the amount of cholesterol carried around in LDL but the number of LDL particles circulating, that actually did not change. I think particle count is considered to be a more informative marker of cardiovascular disease risk. I can't say I'm too concerned about the increase in LDL especially in the context of the weight loss and the reductions in glucose levels. I don't know this for a fact but I would not expect this intervention to increase cardiovascular disease risk. That would be my guess.

Dan Pardi - 18:54: Yeah. A quick question about that. Do you know were there any graphs on the LDL? Did you see the LDL rise and the first couple of months and then stabilize or was it a continuous rise that occurred over the course of the year?

Stephan Guyenet -19:05: I don't recall seeing that. It's a good question.

Dan Pardi - 19:08: Yeah, I'll look into that too because one might reflect the nature of the diet so it rises just because we're eating more fat but if it stabilizes after that point then that would be a different story than one that was continuously rising over the course of the year. Yeah, we'll dive into that after.

Stephan Guyenet - 19:22: Yeah, it'll be more concerning if it was rising continuously. Yeah and then another really important aspect of this study is that in large part it was a one year safety study of the ketogenic diet. They were monitoring adverse events, heart attacks, cancer and stuff like that and seeing if any of those could be linked to the diet. I don't know how hard they tried to link adverse events to the diet since they have a very strong incentive not to do so. However, I will noted they did report there were a couple of heart attacks, there were a couple of cases of cancer in the Virta Health group but we're talking about 349 well no, half of 349 people, however much that is. There were a lot of people. This is not a high level of adverse events given the number of people that were involved.

20:11: I think this certainly is a step in the right direction toward demonstrating that ketogenic diets are safe over periods of a year, they're not going to make you drop dead. They're not going to make your heart explode. They're not going to give you cancer and that's important. I think that's really important because we actually have many people even fairly influential people in the nutrition community who are claiming that these diets will make you drop dead and I don't think that is really consistent with the evidence we have. I think the study adds to that. It seems to be fairly safe. We don't about the long term but over a period of one year, there's not really any indication that it's unsafe. That's important.

20:53: Then one last thing I wanted to mention is adherence. Adherence was great at first. Peoples' ketone levels went way up in their blood indicating that they were adhering to the diet but that really declined substantially over the course of the year as judged by declining ketone levels. They have a really nice graph in the paper. Their ketone levels were definitely still elevated by the end but much less than they were at the beginning. This shows that the average person in the study really was not maintaining consistent compliance to this diet and lifestyle intervention over time.

21:30: This is a really important point because we're talking about a frankly extreme diet. We're talking about a diet that makes radical changes to what you put into your mouth and one that is challenging to adhere to when everyone around you is not eating that way and when you don't necessarily even want to be eating that way yourself relative to all the goodies you could be eating like bread and whatever. It's challenging. I think that most people find this really difficult to maintain and I think it's telling that even in this study where everything was set up for people to have maximal levels of motivation to adhere including paying $370 a month to be a part of this, even then the adherence was not really that great by the end of it.

22:16: We see that the adherence in this study was actually better than in other ketogenic diet studies. In other ketogenic diet studies, most people were really not maintaining the ketogenic state by the end of six months to a year. I think this is really important. We see these evidently positive impacts but I think we really have to ask the question of how realistic it is to ask the average person to eat a diet like this on a continuous basis even if it might be helpful for them to do so.

22:46: One other thing I want to really make very clear is that this study showed that the ketogenic diet as part of this Virta Health intervention controls blood glucose levels relatively effectively relative to other strategies and it's one of the more effective diet and lifestyle strategies for doing so and it caused weight loss and some other associated improvements but that's what it showed. It did not show that these people were no longer diabetic.

Dan Pardi - 23:14: Yeah, that's important. How have these findings been portrayed to the public then?

Stephan Guyenet - 23:18: I think this is really important because if you go on to Virta Health's website it claims that this intervention reverses diabetes and reversed diabetes in the data and this really makes me cringe every time I read it because it's so misleading. The study itself did not claim that the intervention reversed diabetes. I seriously doubt that would have made it pass through reviewers because it didn't even test that hypothesis. They did not even test whether the people in the intervention had diabetes any more by the end.

23:50: I want to be very clear about this. If you're having high blood glucose every time you eat a meal that contains carbohydrate and you stopped eating carbohydrate, you're certainly going to improve your blood glucose levels, but does that mean that you no longer have the underlying pathophysiology of diabetes? Does that mean that you can then go back to eating carbohydrate after that and be fine? No, it doesn't and that's what reversal means.

24:15: If you ask somebody what reversal means, what's the common understanding of that? It's that you no longer have diabetes, right? It's not that you can maintain control over your blood glucose by eating a special diet. That's not reversal.

24:29: What they would have had to do to show reversal and this is presumably why they needed claim reversal in the actual study is to take people at the end and challenge them with carbohydrate and say, "Hey, we're gonna put you on a normal carbohydrate diet for maybe a week just to let your body readapt and then we're gonna challenge you with a standardized sugary drink, standardized glucose drink that they use in research settings and we're going to see how high your glucose goes."

24:56: If their blood sugar didn't spike up to dangerous levels when they drink that then you would say, "Yeah, wow this really reversed diabetes." That's not what they showed and I think this is really important because it's been portrayed by Virta Health and obviously they have a huge financial conflict here. They're essentially promoting this message not for scientific reasons but for business reasons that obviously it looks good to be able to say you can reverse diabetes but what does that say to somebody who starts trying that intervention and thinks that it's going to reverse their disease and thinks probably that after doing this, they'll be able to go back to eating how they were eating and not have diabetic blood sugars? That's what that implies and that's not what they showed.

25:39: I find that really distasteful.

Dan Pardi - 25:41: That's analogous to saying insulin reverses diabetes when it's really just what you can say is that it manages the diabetes.

Stephan Guyenet - 25:47: Exactly, it manages diabetes and furthermore, another really important thing I want to make a point about is that diabetes is more than just about blood sugar. Diabetes is a condition of insufficient insulin signaling, whether it's a combination of your tissues not responding to insulin very well and your pancreas no longer being able to produce enough insulin to overcome that and what happens then insulin plays a really important role in controlling your blood sugar levels. When it's not working right, you get really high blood sugar but that's not the only thing insulin does. You get all kinds of dysfunction in all kinds of body systems when insulin is not working right. Insulin also controls your fatty acid levels in your blood, high fatty acid levels are harmful as well, and it controls many other physiological processes.

26:36: Glucose is the thing that we test to see whether you have diabetes or not but that is not the entire disease. That is a marker of the disease and it's an important marker. It's probably one that plays an important role in the actual bad outcomes of the disease like heart attacks, blindness, and peripheral nerve damage and kidney damage but is that really the only thing such that if you can get your blood glucose under control you're fine? The answer is we don't know. We really don't know.

27:06: I think it's really important to understand is that this Virta Health study they showed that it controlled blood glucose fairly effectively and that's great. That's a good thing but we actually don't know yet whether that style of diet and lifestyle will be able to prevent the things that we really care about preventing in diabetes, the things that really matter like whether you're going to go blind or not and whether you're going to have to be on dialysis or not and whether you're going to have a heart attack or stroke. We don't have evidence of that yet. I would guess that it would be beneficial based on things that I've seen but right now we're making guesses based on indirect evidence not direct evidence. This study did not provide direct evidence of that.

Dan Pardi - 27:47: Yeah, super important and you mentioned that earlier that one of the results was that it cut insulin in half and that could be reflective of an improvement of the disease but it's not indicative of curing it otherwise people would have gotten off the insulin altogether.

Stephan Guyenet - 27:58: It's possible that some of the people who went off insulin would not be diabetic anymore if they went back to a normal diet. I don't know. We just don't know that I will note that the main use of insulin that people use is to cover carbohydrate in their meals. If you're not eating carbohydrate, you're not using insulin to cover the carbohydrate in your meals obviously you're going to be using less insulin. That doesn't necessarily reflect change in the underlying pathophysiology of the disease.

28:31: Another public comment that I want to comment on because I think it really teases out some important issues. When the study came out, obviously low carbohydrate diet advocates were very excited about it. Virta Health was very excited about it and there was an editorial in the LA Times by Gary Taubes and Nina Teicholz I don't know if I'm pronouncing that correctly but I want to pull one quote in this because I think it dredges up some interesting scientific issues. They said, "Early results of a current trial reported that type 2 diabetes symptoms can be reversed in just 10 weeks on such a diet. They effectively cured themselves of their disease something that mainstream medicine does not even believe possible."

29:15: We've already discussed the idea of cure versus management so obviously that part is incorrect but I think there's another issue here that's really important for listeners to understand and that is the other part of their statement where they say that this is something that mainstream medicine does not even believe is possible to cure diabetes in fact that's not correct. Not only does mainstream medicine believe it's possible but mainstream medicine has already come up with at least three different ways to do it and unlike in the Virta Health study these ways have actually been validated to actually put diabetes in remission so that people can eat carbohydrate again not just control glucose.

29:58: I'm just going to talk a little bit about what those are here. Just last year in The Lancet, there was a large and well conducted randomized controlled trial showing that a very low calorie for three to five months in people who are pretty similar to the people in the Virta Health study reverses diabetes in nearly half of people and they're still not diabetic a year later. As in, they can eat carbohydrate again and eat a normal diet again without dangerously high blood sugar. If you look at the people who didn't achieve remission, you can see that they didn't lose much weight suggesting that they didn't actually adhere to the low calorie diet. Among those who lost an amount of weight consistent with sticking to the low calorie diet, 86% of them reversed their diabetes. This is extraordinary and this is mainstream medicine we're talking about. Obviously, not only do they believe this is possible but they have already demonstrated that it is possible just not using a ketogenic diet.

30:58: Another way that mainstream medicine has reversed diabetes is through bariatric surgery, weight loss surgery. Meta-analysis suggest that 77% of people with diabetes no longer have diabetes after a bariatric surgery. Again, they can actually eat carbohydrate again without having diabetic blood sugars, 77%. That's a huge very powerful effect.

31:25: The third thing that mainstream medicine has come up with that reverses diabetes is early insulin administration. As soon as diabetes develops, if you get people insulin right away I don't know exactly how it works. Maybe it just gives your pancreas a rest for a little while but if you give them insulin right away a certain proportion of people no longer have diabetes. It actually short circuits it at least I don't know how long the effect lasts but it does seem to at least temporarily reverse the diabetes.

31:59: That's three different ways that mainstream medicine has actually shown how to reverse diabetes, which the Virta Health study did not show. Not to say that it doesn't, it might actually reverse diabetes and I would love to see data on that. I think they should test that but currently we just don't know. I would guess that it probably did at least in some people due to the impressive weight loss.

32:22: Then there's one last point I want to make and I think this is just as important if not more important is that not only has mainstream medical research developed effective ways to reverse diabetes, it's also developed a very effective way to prevent it in the first place.

32:37: We have several large, multi-year randomized controlled trials showing that modest weight loss plus exercise, the type of diet doesn't seem to matter reduces the risk of developing diabetes by about 60%. These trials were measuring actual diabetes outcomes not biomarkers. They were actually measuring the development of actual diabetes not just a biomarker. These studies showed very conclusively and again high quality randomized controlled trials in multiple populations on multiple continents, multiple different races and cultures showing the exact same thing that with weight loss and exercise, even if it's a low fat diet, even if it's a portion control diet, you can get impressive reductions in diabetes risk and the greatest predictor of the reduction and risk is the amount of weight loss. The weight loss really seems to be the main key.

33:33: One of the reasons why this is so important is that we don't have this kind of evidence for really any other chronic metabolic disease. There's no randomized controlled trial showing that an intervention can reduce your risk of having a heart attach by 60%. That's absurd. We've really actually have a very good idea of what diabetes is, how it happens, and how to prevent it and even how to reverse it. It really seems to come down to energy balance. How many calories are you taking in? How many calories are you expending and how large are the fat stores on your body?

34:09: It's very plausible that Virta Health and a ketogenic diet could achieve something similar. We don't know yet but I would love to see more research on it.

Dan Pardi - 34:18: Yeah. Those are some really important points. We have potentially another effective technique to combat diabetes. We don't know that it reverses it but it could help to manage it and more research is needed in order to know can people actually tolerate glucose again in their diet. That's not been done here but it has been shown previously with other techniques.

34:38: Now I would imagine that some ketones are made on a very low calorie diet, which was one of the ones that had a huge reversal in The Lancet. I'm curious is it just the fact that people can maintain a calorie deficit more easily, which then facilitated the weight loss or are there some unique properties of ketones themselves that are both affecting weight loss and maybe separately diabetes and pancreatic health? That remains to be seen.

Stephan Guyenet - 35:01: Yeah, it's a good question and there's a lot of research that's been happening on this. Nothing really definitive that I've seen yet but yeah, there's something interesting going on here.

35:11: One of the things we've talked about a lot on your podcast is the setpoint. This is something I focus on a lot into my writing and into my book. This is the idea that your brain is defending a certain level of body fatness and that's what makes weight loss so hard and may even be part of what drives obesity in the first place. I'm always on the look out for interventions that can lower the setpoint. Basically, make weight loss and maintenance more comfortable and sustainable by making the brain "want" to be at a lower body fat level on a non-conscious level of course. That's what the idea of weight program to a large extent is based around, trying to lower your setpoint using different strategies that seem to do that.

35:55: When we look at the ketogenic diet, we see that people spontaneously reduce their calorie intake by a substantial amount and that they're not trying to reduce their calorie intake, they're just spontaneously eating less and that to me suggests that their setpoint may well be lowered. That's why it's easier and more sustainable for them.

36:15: I think that's another interesting aspect of this and for me if I had type 2 diabetes, there's still a lot of uncertainty about the long term effects of the ketogenic diet but if I had type 2 diabetes and was having trouble controlling my blood sugar, I would consider doing something like this. Especially, if I had $370 a month burning a hole in my pocket.

Dan Pardi - 36:35: Right and we've discussed this before. I have experimented with the ketogenic diet. I have even worked with other friends who have done it and I've noticed some pretty interesting things. When somebody first goes on to a ketogenic diet particularly for the first time, it's not uncommon while people will experience a lot of weight loss fairly rapidly in the maybe 10 to 20 pounds within a short amount of time. Some that's probably water because they're taking in less carbs but some leanness that occurs just by being on that diet that seems unique.

37:02: When people do it a second time however, the results are a lot less impressive. That's commonly reported and I know personally for me when I'm eating three square meals a day in ketogenic mode then my results really plateau. For me to get good adherence or as you get good results in terms of promoting leanness, it's usually like I'm eating two meals and a snack and it's a calorie deficit. I think that there could be some unique benefits to just me helping you with calorie control but then there's a behavior component, which then is, "Okay, don't just be eating food because it's in front of you". That's been educational.

Stephan Guyenet - 37:34: Yeah, that's interesting. I went through a period of low carbohydrate eating years ago and it wasn't ketogenic but I was eating a lot less carbohydrate than an average person. I noticed that it really smoothed out my hunger pangs. It's not that I was eating fewer calories, it's not that my weight was changing but I didn't have as acute motivation to eat around meal time. It was easier for me to skip meals. It was easier for me to skip an entire day of eating. I would fast. I would do 24 hour water fasts every week and it's a lot harder for me to do that now. It's not like I'm not motivated enough to make it happen but it makes my brain fuzzy and I can't work effectively if my brain's fuzzy because I have pretty brain intensive job. I did think that was interesting how it smoothed out my food motivation to where it wasn't as acute around meal times. I could definitely see how that would help someone with making positive dietary changes.

Dan Pardi - 38:34: That's another interesting component. Any diet you talk about whether it's the Mediterranean diet, low carb, high carb, keto, what is the quality of that diet? You could have two versions of the exact same diet and they might be similar in a macro nutrient profile but they're actually quite different in terms of the nutrients that you're getting in your body. I think for a lot of people the ketogenic diet helps them rule out a huge class of dangerous foods but then it could also rule in things that we know are also risky so eating a lot of salami and perhaps cheeses. You have a diet that's just peeling back the skin on a salami stick all day. That's your diet, right?

39:10: For me, I noticed that modifications when I do ketosis a couple of times a year and I do it for different reasons than just some weight loss but there's some aging benefits as you've heard with my podcast with Neil Copes potentially that I'm curious about. I actually eat more vegetables probably on a ketogenic diet so non-starchy greens with a little bit of meat and lots of olive oil. It's a highly Mediterranean version. I actually don't measure my ketones but it's almost a heuristic that guides me towards really eating a moderate amount of meat, lots of veggies, and a little bit of fruit. It's a version that I do like.

Stephan Guyenet - 39:39: Yeah and I hope that, that's how other people implement the ketogenic diet. Obviously, it's a diet that can include a lot of whole plant foods, a lot of vegetables and probably can be nutritionally adequate but it really depends on how you implement it.

Dan Pardi - 39:54: Yeah, definitely well we could go down the rabbit hole of ketosis and ketones another time but I think keeping it focused on what this study did, what it found, who was it in, some of the caveats that again like you said it made you more skeptical to take a closer look but that's not cynical. Cynical means you're going to disregard the results because of the connection to the corporation that's conducting it. It's just like, "Okay, since this corporation is conducting it that has a financial stake in its outcome, we're just going to look at it closely," and I think you've brought up some really important points about the ways that it's been misinterpreted in the literature by people that are promoting low carb regularly.

40:28: My summary of it and correct me if I'm wrong but is that this is all conducted. There are some limitations. It's exciting results but more research is needed because they're a really important outcome of being able to handle glucose as a marker of the state of pathology of diabetes. That has not been done and so we need more research on that to make further claim. Those results could have been there but we just don't know because they weren't tested and reported on.

Stephan Guyenet - 40:52: Yeah, I agree with that. Overall, this study I found it to be pretty clean and what I really object to is how it's been portrayed after the fact but the study itself, well written, well conducted, as you said it has limitations but I think despite those limitations, it provides a valuable update.

Dan Pardi - 41:10: Well that's probably a good place to end. Stephan, it's always illuminating to have you on. Your shows are always disproportionally popular so kudos to you and thanks again for taking the time to come on the show.

Stephan Guyenet - 41:21: Okay, thanks. My pleasure.

Kendall Kendrick - 41:25: Thanks for listening and come visit us soon at humanOS.me.