[HD: This is the second part of my duo about Diabetes and the Ketogenic Diet. In Part 1, Why This Endocrinologist Hates Diabetes, I discuss…uh…why I hate diabetes. You don’t have to read Part 1 to appreciate Part 2, but I recommend it just for context.]
The available strategies for weight management in type 2 diabetics don’t work very well. It’s not that the advice to eat less and move more is bad, but people just aren’t very good at following it in a way that is sustainable, producing lasting results. On top of that, I give them diabetes medications (e.g. insulin) that may impede their progress. So, I ask: what can we do to control blood sugar and reduce weight, which will ultimately lead to better long-term control of glucose levels by reducing insulin resistance?
Enter the Ketogenic Diet!
Those of you who are familiar with not only my writing style, but also what gets my quack-o-meter dialed up to 11, are probably thinking, “Ooh…I can’t wait to see HD rip KD a new one. Let’s get it on!” As much as I enjoy debunking extreme diets based on lousy/no science (hcg diet, anyone?), I also enjoy presenting ideas that are extreme, but also may have merit. This is one of those times.
So what is a ketogenic diet (KD)? There are plenty of sites where you can dig into the details, so I’ll present just a quick overview. In general, it’s a very low-carbohydrate diet that tends to be high in fat. There are various iterations, but a common macronutrient percentage-of-daily-calories breakdown would be 70/20/10 fat/protein/carb.
70% fat?! That sounds revolting! I’m clicking back over to my Twitter feed now. See ya!
Listen, I’m with you. I don’t have any desire to eat ketogenically. I trim every last bit of fat off my steak before a bite enters my mouth. However, if I was overweight with diabetes and I had already tried “everything,” I think I would be receptive to hearing more. And in fairness to KD, it’s quite possible to do it without chomping on a stick of butter like it’s a Clif Bar. So, for those of you with some residual interest, read on. The rest of you, I hope the Kardashians did something really interesting today.
When should we think about eating a Ketogenic Diet?
First, it has actually been around as a treatment for refractory epilepsy since the 1920s. And it has been used as a treatment for obesity since the 1960s, so it has some longevity. Over the last decade or so, it has been investigated as a tool for helping treat: diabetes, PCOS, neurological/respiratory/cardiovascular diseases, cancer, and acne.
Um, the claims for KD kind of sound like the claims that have been made for low-dose naltrexone – it seems to treat everything. That’s raising red flags on my quack-o-meter.
Ah, young Jedi, I have taught you well. You should be skeptical about anything that purports to treat everything, and that includes KD. I am not saying that KD is the magical answer to all of our medical problems, but there is some compelling science that suggests it may be able to back up some of its seemingly fantastical claims.
What happens to our physiology when we eat a Ketogenic Diet?
After a few days of fasting or eating a very low-carb diet (<50 grams/day), glucose reserves become insufficient for normal fat oxidation via the Krebs cycle. Remember the Krebs cycle from Intro to Biochemistry? Neither do I. Glucose reserves are also insufficient for the central nervous system (CNS) to use as fuel. Since the CNS doesn’t have access to sugar, and it can’t use fat for energy, it is forced to find alternatives.
At this point, the body is overproducing acetyl coenzyme A, which I’m only mentioning because this leads to increased production of ketone bodies (KB). This is called ketogenesis and occurs in the mitochondrial matrix of the liver.
Under normal conditions, the concentration of KB in the blood is < 0.3 mmol/L, while glucose is present at 4 mmol/L. When KB increase to around 4 mmol/L, the brain will start to use them as fuel.
Glucose levels in the blood will be in the low-normal range, maintained by glucogenic amino acids and from glycerol liberated by lysis from triglycerides. Don’t worry if that sentence sounded like gibberish; I just want to get the point across that the ketogenic diet does not result in hypoglycemia (low blood sugar), as long as you’re not taking medications that lower blood sugar. If you are taking such medications, they will almost certainly need to be adjusted down or stopped on KD.
Does the Ketogenic Diet produce weight loss?
Before I answer that, let me say that I dislike the question, “What is the best diet for weight loss?” First, I hate the word “diet,” as it implies a temporary change in eating habits – usually through draconian calorie restriction – which is not a solution to the chronic problem of obesity. Anyone who has ever dieted knows that cutting calories drastically will result in weight loss. They are also painfully aware that the return to a more sustainable intake is associated with regain of all the lost poundage (kilogramage for our European and other metrically-inclined friends) and then some.
To reframe the question, I would ask, “Is the ketogenic style of eating typically associated with weight loss?” The answer is, it depends (like almost everything in medicine). Remember, you’re now getting somewhere around 70% of your daily calories from fat. Since fat is the most calorically dense nutrient (9 calories/gram, compared to 4 calories/gram for protein and carbs), you have to be careful when loading up the diet with fat. It would be easy to gain weight on KD by eating so much fat that your daily calorie intake increases above baseline.
However, because you will take in a fair amount of protein with the fat, and protein turns the brain’s hunger center off, appetite can be suppressed without having to think about counting calories. There is also a suggestion that levels of the appetite-regulating hormones ghrelin and leptin change on KD, leading to greater satiety. It is also possible that there is a direct effect of ketone bodies on appetite suppression, but this is conjecture.
Studies show that these very low-carbohydrate diets have the potential to produce more weight loss than a similar calorie standard diet. Is it all due to decreasing appetite and increasing satiety? Probably not. There are some other key differences between KD and a standard diet:
- Reduced lipogenesis (fat formation) and increased lipolysis (fat breakdown)
- Increased metabolic efficiency in consuming fats
- Increased metabolic costs of gluconeogenesis (glucose synthesis) and the thermic effect of proteins
Does the Ketogenic Diet cure or reverse diabetes?
To answer that question, we have to look at what happens to diabetics’ physiology if they change from a standard diet to KD. People with type 2 diabetes are insulin resistant, diverting a fair amount of dietary carbohydrate to the liver, where it is converted to fat, as opposed to being oxidized for energy in skeletal muscle. By switching to KD, dietary carbohydrate is restricted to a level so low that it does not get converted to fat. Remember that more visceral body fat = more insulin resistance = higher blood sugars. So, if we can decrease visceral fat by restricting dietary carbohydrate, we can decrease insulin resistance, thereby lowering blood sugars.
In addition, studies show that higher levels of ketone bodies (3 mmol/L) are associated with lower glucose output from the liver, another win for the blood sugar level.
Interestingly, these changes in glycemic control and the necessary reduction of diabetes medications happen before significant weight loss has occurred, like what we see immediately after gastric bypass surgery. This would suggest that there are other factors at play (as mentioned above) in the improvement in diabetic control, making the point that sugar intake does correlate with diabetes risk, independent of weight or sedentary lifestyle.
Stop teasing me with all this mental masturbation and give me the bottom line! Does eating a Ketogenic Diet cure or reverse diabetes?!
Gosh, you’re impatient. I promise I will answer this, but let’s first look at some data. A 2008 study compared the effects of a low-carb, ketogenic diet vs. a low-glycemic index diet on glycemic control in type 2 diabetes (Westman EC et al. Nutr Metab [Lond]). The LCKD showed better A1c (3-month average blood sugar) reduction, weight loss, and increased HDL over the 6 month study period. Diabetes medications were reduced or eliminated in 95% of the LCKD group vs. 62% of the LGID group. Sounds promising so far, no?
A 2017 study compared a LCKD vs. a standard, ADA “plate method diet” in overweight people with type 2 diabetes (Saslow LR et al. JMIR). The results were better in the LCKD group with respect to weight, A1c, and triglycerides. However, the intervention group got way more attention than the control group, with mindful eating coaching; plus, the control group diet wasn’t novel, probably explaining why the dropout rate was 50% compared to 8% of the LCKD group. With small numbers to start with, that study did not impress me.
To muddy the waters a bit more, I’m going to throw in a 2014 22-week study in mice (mice data ≠ human data, but bear with me), which showed that LCKD caused glucose intolerance, reduced pancreatic beta- and alpha-cell mass, higher cholesterol and triglycerides, and did not result in weight loss (Ellenbroek JH et al. Am J Physiol Endocrinol Metab). It’s possible that the type of fats (high saturated, low polyunsaturated) served to the mice may have made things look way worse in this study. Perhaps results would have been different had they served more polyunsaturated fats with omega-3’s.
I’m obviously not performing a meta-analysis here, but what I’ve presented was fairly representative of the body of literature out there on this subject, until now…
Enter Virta Health
Virta Health is, according to their own website, “an online specialty medical clinic that reverses type 2 diabetes without medications or surgery.” [HD: I have no financial relationship with Virta Health, though I would welcome them as an advertiser on this site if they were so inclined. Relax, I’m not going to smooch their behinds just to make a few bucks; I’ll give you my honest impression of their operation.]
So how are these guys any different from your local weight-loss shop that offers the ketogenic diet? You know – the one in the strip mall with the “Lose Weight Here!” banner hastily tacked up and listing to port, pasted over the old sign for Deb’s Dry Cleaner. Well, first off, they’re approaching this problem with some degree of scientific rigor, and they score major points for that.
Virta partnered with Indiana University and published a paper in 2017, detailing their 10-week results from a trial consisting of 262 patients. As you can imagine, the results were favorable with respect to A1c reductions, weight loss, and medication reduction/elimination. I’ve already tipped my hand and made it clear that I think what they’re doing has merit, so let’s take a moment to cover the flip side of the coin – weaknesses.
First, the paper was published in JMIR (Journal of Medical Internet Research). Despite the fact that it has a very respectable impact factor of 5.175 (just below my beloved Journal of Clinical Endocrinology and Metabolism at 5.455), I don’t have as good a handle on how rigorous the peer-review process is for submissions to the journal. At least there is a peer-review process, and revisions must be completed prior to publication. But, given JMIR’s mission to fast-track publications and get information into our hands faster than traditional print journals, one has to wonder if the quality of the studies accepted is on par with JCEM, Thyroid, and other top journals in Endocrinology.
Second, though the study had a robust number of participants, it was very short, there was no similar-calorie diet control arm, and the patients in the study were not representative of my patient population (average A1c at entry was 7.6, compared to my patients in the 9-12 range). The question is whether patients who have had diabetes for longer durations – or have much worse control at baseline – can achieve results similar to the study population. Remember that such patients may have significantly decreased pancreatic reserve, such that they will need diabetes medications regardless of what they eat and how much weight they lose.
Virta has recently released but not published their 6-month data, and I will commend them if they continue to release data every 6-12 months for the entire, planned, 2-year study duration. Their 6-month data shows 89% retention of study participants, which is a testament to the program, as most people don’t last that long in other commercially available weight management programs. Inexplicably, they allowed subjects to decide whether or not to test their A1c’s at the 6 month mark, so the average A1c of 6.1 among those 108 people should have a huge asterisk next to it – did the people who knew they were not doing as well refuse the A1c test?
They have robust data for weight, though; their subjects lost an average of 12% of body weight, with 81% of people losing at least 5%. Given that we usually don’t get more than 5-10% weight loss with appetite suppressant drugs like phentermine, this is impressive, especially if it can be maintained.
What makes Virta Health special?
Well, they’ve set up a comprehensive program to tackle an intractable problem that nobody else has managed to figure out a solution to and scale that solution. Sure, you probably have some local, academic institution with an incredible diabetes and obesity program, but that doesn’t do the rest of the country any good. Plus, patients want remote care nowadays. I typically counsel at least one or two people daily to go to some sort of formal weight management program. You know how many follow through with that? Less than 5%. Why? Most common answer: “I don’t have the time.”
Virta has physicians and health coaches continuously guiding their clients, remotely. Unlike in-person doctor visits that may take place once every few months, the Virta team has touch points as frequently as the needs of the client dictate. Having a hard time figuring out what to eat for breakfast because you’re bored with what you’ve been doing? Talk to the health coach. Feeling the “keto flu” in the early stages of the program? Chat with the physician and health coach. Down because you feel like you’ve stalled with your progress? Get a pep talk and some pointers for breaking through that plateau from the health coach. If it sounds like I’m enthusiastic about this type of program, it’s because I am. I could never, in a million years, offer this level of service to my patients. But I believe it is what many of them need to be successful and to sustain that success.
Would most diabetic patients do significantly better with a personal health coach and frequent check-ins with their physician? Probably. Do they need to follow an extreme diet, or is it the intensive followup that drives the results? Long-term, I suspect that the intensive followup will prove to be the most important part. But in the short-term, it’s clear that KD causes blood sugars to drop dramatically, even before much weight loss has occurred. As I said earlier, I would have liked a similar calorie, standard-diet control group so we could really gauge over time the relative importance of the dietary interventions vs. the intensive followup.
You’ve managed to get almost 2500 words into this thing without answering the question you promised you’d answer…
Yes, I’m aware of that; it’s called building suspense. Or maybe I’m just too verbose and I get off topic; I’ll shift gears now. If you want to hear more about Virta Health, listen to one of their doctors, Jeff Stanley, on Dr. Scher’s Boundless Health podcast (Episode BH011).
Does the Ketogenic Diet cure or reverse diabetes?
In my opinion, I think it is overly optimistic and slightly misleading to say that KD cures or reverses diabetes. I would prefer that Virta Health and other proponents of KD use terms like “diabetes in remission” or “diet-controlled diabetes.” Here’s why:
In order to label someone “cured” of type 2 diabetes, I think that person should meet two criteria: she should have normal glucose tolerance/no insulin resistance, and she should have pancreatic reserve capable of disposing of any glucose load to the system. Let me explain. By the time someone is diagnosed with type 2 diabetes, she has been insulin resistant and has had impaired glucose tolerance for years. She has already lost up to 50% of her beta cell (insulin-producing cells of the pancreas) function by the time the blood sugars are high enough to make the diagnosis of type 2 diabetes. As time goes on, beta cell function continues to decline, until eventually the patient makes so little insulin that non-insulin drugs are no longer sufficient to control the disease, and insulin replacement therapy is needed.
In order to be “cured” of diabetes, beta cell function should be sufficient to handle any glucose load that comes in to the body. If beta cell function is still fairly low while on KD, then the diabetes is really just in remission or diet-controlled for as long as that person is eating a ketogenic diet. If that person falls off the ketogenic wagon, we can assume that she will either immediately have poor glucose tolerance, or she will soon develop high blood sugars as her physiology goes back to her prior abnormal state and her weight increases.
In my estimation, this does not represent a reversal or cure of the disease. It certainly represents an impressive remission, just as gastric bypass surgery can induce remission in type 2 diabetics. But what do we tend to see over the next decade or two after bypass surgery? You guessed it. Patients begin to regain their weight, and the diabetes returns with the weight. Cured? Not so much.
The Ketogenic Diet might reverse diabetes.
This is what I love about medicine. Just when you think you’ve made a cogent argument and you’re ready to smugly take your seat, somebody comes out with a study that challenges that argument. In the February 2017 issue of the journal Cell, a study was carried out by researchers from USC, MIT, and some place in Italy of which I’ve never heard. But if they partnered with USC and MIT, we should probably be impressed.
In the first part of this study, they looked at mice with type 1 diabetes, mice with type 2, and normal mice. They put the mice on what they call a fasting regimen, which was basically a ketogenic diet. At the end of each 4-day cycle of KD, the mice were fed regularly for up to 10 days to ensure they regained their body weight, before beginning the next fasting cycle. They went through 3 cycles of fasting/regular feeding.
In the type 2 diabetic mice, after the fasting cycles, insulin secretion was restored and insulin resistance was reduced. The researchers think that beta cells were regenerated. In the type 1 diabetic mice, the fasting cycles resulted in an increase in the number of beta cells generating insulin (remember that type 1 diabetics should make little to no insulin).
In the second part of this study, they recruited healthy, non-diabetic, human volunteers, who underwent three cycles of a similar fasting (KD) regimen. They then took blood from these healthy people and applied it to cultured pancreatic cells from humans with type 1 diabetes. The results seemed to suggest that it is possible to reprogram cell lineages and generate insulin in these diseased pancreatic islet cells.
Is this overwhelming evidence that the Ketogenic Diet reverses diabetes? No, of course not. Over the years, we’ve seen several promising diabetes “cures” come and go. The Cell study is interesting, for sure, but mouse data is notoriously difficult to extrapolate to humans, for one. Secondly, the human data from that study was basically in vitro data, so we really don’t know if that effect would translate into clinically meaningful remission from diabetes in a real human.
Type 2 diabetes and obesity are two extraordinarily vexing problems that we have mainly failed at controlling with lifestyle modifications alone. The Ketogenic Diet is an extreme eating strategy that may hold great promise for those who can stick to it long-term. In my opinion, for the majority of people out there, KD is not a do-it-yourself proposition. In my office, the most common phrase I hear with respect to KD is, “I did the Ketogenic Diet…for 2 weeks.” There are just too many ways to fail on this diet, and I believe expert guidance is needed.
I think companies like Virta Health have the right approach, offering a remote, comprehensive program with physician oversight and frequent feedback/coaching. I think their early data is promising, and I’m excited to see if they can show sustained remission from diabetes over the course of years. If they can, it will be a game-changer in the Kingdom of Endocrinology.
Are you a doctor who treats diabetes or a person living with diabetes? Have you found success with the Ketogenic Diet? Do you think it’s extreme and difficult, or did you/your patients acclimate pretty quickly? Would you refer your patients to a company like Virta? Why or why not? If you have diabetes, do you think a program like Virta’s would be something you could stick with, assuming it’s affordable? Comment below!
By reading this site and interacting with me in the Comments, you agree to abide by my Disclaimer.