Thoughts on the Pritikin Diet

Denise Minger wrote a very thought provoking post on the very low fat Pritikin Diet and how it can help reduce disease including diabetes and cancer. In the first part, I wrote my thoughts on the Kempner Rice diet here. The Pritikin diet is primarily composed of high fibre, high carbohydrate, low-fat, plant based foods. With his close relationship to McGovern, this diet formed the basis of many of the low-fat, carbohydrate based recommendations of the 1977 Dietary Guidelines for Americans.

Over some decades, his colleague RJ Barnard published many studies on this diet and how it could reduce many different diseases – mostly those associated with the metabolic syndrome. For example, this diet could help reduce prostate cancer, reduce insulin and medication usage for type 2 diabetes, reduce LDL oxidation, reduce breast cancer, reduce colon cancer, reduce heart disease, and improve cardiac risk factors.Minger1

Furthermore, there are many traditional societies that ate a similarly ultra low fat, whole, unprocessed vegetable based diet that had low incidence of all these diseases. The Okinawans and the Kitavans come to mind quickly.These are many of the same benefits that low carbohydrate diets also claim. How can very low fat diets have the same benefits as very low carbohydrate diets? Perhaps there are, as Denise suggest, two different kinds of ‘magic’ going on?

The Pritikin results are clearly at odds with the Carbohydrate-Insulin Hypothesis (CIH) which holds that high carbohydrate diets result in high insulin and therefore obesity, T2D, and the rest of the metabolic syndrome. High carbohydrate diets should have the opposite effect as high fat diets.

So, is Denise right? Does the Pritikin diet have some ‘magic’? Here’s a surprise. I almost completely agree with Denise.

The answer is this. Macronutrients, just as with calories is the wrong measure of a diet. This is nutritionism’s greatest blunder. This refers to the fact that trying to reduce the complexity of foods and their effect on humans to some mix of macronutrients (fat, protein and carbs) is completely wrong.


The CIH, depicted above, as it stands is clearly inadequate. There are too many inconsistencies, which I’ve written about in a previous post – unimaginatively titled “The Carbohydrate-Insulin Hypothesis is Wrong“. The Asian diet of the 1990s for example were very high in carbohydrates, but did not lead to obesity. The main problem with the CIH is that it does not take into account the multiple factors that lead to an increased insulin level (you can review my 50ish posts of the Calories/ Hormonal Obesity series for more background).

Increased insulin leads to obesity and the metabolic syndrome. And sure, refined carbs lead to increased insulin, but they are not the only things that do so. This sort of simplistic theory does not account for, among other things, the deadly effect of fructose, the time dependency of insulin resistance, animal protein, fibre, vinegar and fermentation effects, or dietary fat. CIH pretends that the only thing that matters is the amount of carbohydrates. So it’s wrong. But it’s not really incorrect, so much as incomplete. Yes, carbohydrates are a big part of increasing insulin, but they are not the only part.


A more complete version would look like this – the Hormonal Theory of Obesity. Notice that insulin is still at the core, but there are many inputs and protective factors – only one of which is carbohydrates. There are many things that raise insulin and many things that also lower insulin. The CIH was too simplistic, and therefore, incorrect.GL

In addition, the use of macronutrients instead of foods complicates things significantly. Not all carbohydrates are the same, just as not all fat or proteins are the same. For example, the amylopectin (a type of carbohydrate) produces vastly different effects depending on whether it comes from wheat or from beans.  Amylopectin A is not the same as Amylopectin C. You can see this effect in the Glycemic Index.

Carbohydrates are not all equal. Unprocessed, whole carbohydrates, yes even boiled potatoes have a much, much smaller effect on glycemic load than processed carbs like cornflakes. For carbohydrate foods, there is a close correlation between GL and insulin secretion. So 30 grams of carbohydrates can either be the most healthy looking lunch you’ve ever seen, or 1 hamburger bun. So focusing simply on macronutrients, whether fat, protein, or carbs is simply wrong. We need to talk about foods, not macronutrients.Slide4

So, looking at the more complete Hormonal Obesity Theory (HOT), you can see that insulin is what drives obesity, but carbohydrates are not necessarily what drives insulin. Non-caloric, non carbohydrate factors such as cortisol (stress) can also play a big role in insulin and weight gain.

So the key question with regards to the Pritikin Diet is not how many carbs, or how much fat, or how many calories. The hormone insulin is what drives obesity. Therefore, the key question in this diet, as with all diets is this:

What happens to insulin?

If a high fibre, high unrefined carbohydrate, diet lowers insulin, then we should expect all the benefits of reduced weight and reduction in all of the manifestations of hyperinsulinemia – Type 2 Diabetes, Hypertension, heart disease etc. If a Low Carb High Fat diet also lowers insulin, then we should see all the same results. And guess what? I don’t really care if you eat LCHF or Pritikin – as long as those insulin levels come down, I don’t care. Why? Because you will derive the same health benefits of lowering insulin. This will be highly beneficial in all the diseases of high insulin (metabolic syndrome).barnard_2008_graph

In other words, we need to evolve from a carbohydrate centric view of obesity (CIH) to an insulin-centric view of obesity (HOT).

So let’s look closer at the Pritikin diet, and others based on it, like the Ma-Pi diet. Insulin goes down. A lot. Even though carbohydrate consumption is high.

Why? Well, the Hormonal Obesity Theory explains why. The carbohydrates being eaten are whole unprocessed ones which do not increase insulin as much (low glycemic load). There is little to no fructose. There is lots of fibre (protective). There is less animal protein (insulinogenic).

This is not a high insulin diet! It lowers insulin. Therefore, I would expect that it lowers weight and the rest of the metabolic syndrome as well. And that’s exactly what Pritikin, as well as Neal Barnard and others have shown. Obesity, T2D, and the rest of the metabolic syndrome are simply different manifestations of hyperinsulinemia (too much insulin). Therefore, any diet that lowers insulin will show incredible benefits. This includes the Pritikin Diet as well as the LCHF and Paleo and Atkins diets. Forget about macronutrient composition of diets. I’ve always tried to stress one point – The toxicity lies in the processing, not the macronutrient composition. You could even live in the macronutrient swampland and do well. But you must keep insulin levels low.Kitava-Insulin

So, I do disagree on one semantic point with Denise. I don’t think there are two types of ‘magic’ here. There is only one type of ‘magic’ – lowering insulin results in benefits in a high insulin disease state. But there are two different ways to get there. Both Pritikin and LCHF work fine. Actually – there are multiple ways to get there – Mediterranean, potatoe diet, Zone Diet, Weight Watchers etc.

Can you eat a high carb diet and still have low insulin? Of course. Because carbohydrates are not the only stimulus to insulin. The Kitavans – a traditional Pacific Island society demonstrate this perfectly. Eating a diet that is estimated at 70% carbohydrate – unprocessed, whole foods with almost no sugar, the Kitavans maintained a serum insulin level lower than 90% of a typical Swedish person.

So, you might be wondering, why do I describe my diet as predominantly LCHF with Intermittent Fasting (another great way to lower insulin)? Well, because most people want the Twitter version of my dietary theories, rather than slog their way through 50 or 60 posts (although my upcoming book will make that easier). The closest approximation is LCHF with IF.  It also differentiates it with the failed Low-Fat movement of the last 50 years.

One last point. I absolutely love the fact that Denise entitled her post – A Call for some Evolution of Thought. Because too often we are bogged down in the Low-Fat vs Low-Carb wars. That’s soooo 2007. Or Caloric Reduction as Primary (CRaP). That’s soooo 1982.Or the CIH. That’s soooo 2010. We need to evolve our understanding of what causes obesity. What is the aetiology of obesity? (That was the original title of my book – The Aetiology of Obesity, but the editors hated it. Too nerdy)Imbalance

We need to move forward from the failed ‘Calories causes Obesity’ paradigm. But just as surely, we need to move forward from the failed ‘Carbohydrates causes Obesity’ paradigm. Let’s face the facts head on. We’ve done Atkins. It was not the effortless weight loss promised. The underlying Carbohydrate-Insulin Hypothesis was woefully incomplete. But the Insulin Hypothesis is still intact. Obesity is a hormonal, not a caloric imbalance. But carbohydrates was not the entire story. There is so much more – stress, and sleep, and cortisol, and protein, and fibre, and fat, and fructose, and insulin resistance and vinegar, and fermentation, and the incretin effect.

Forget macronutrient content. It’s not the Fat. It’s not the Carbs. It’s not the Calories. As my friend Amy Berger of TuitNutrition says, It’s the Insulin, stupid!

45 Responses

  1. Don’t know where you got your numbers, but
    states that glycemic load of white bread is 10 and
    GL of grapes is 11.

    Glycemic load is too subjective. Who knows
    how much a person eats of any food item? The
    better gauge of food’s effect is glycemic index.

    A much better analysis is Marty Kendall’s site:

    • Without re-working the math, I would guess that Dr. Fung adjusted the load data according to a constant weight of the items he graphed, instead of the “serving size” as reported on the site at the link you posted. So he may be reporting glycemic load per 100 g.
      On the site you link, the serving sizes vary by quite a bit. E.g., the white bread serving size is 30 g, but the grapes serving size is 120 g.

      • Jane:

        Apparently you have never heard of the “glycemic load.” It is
        the “glycemic index” adjusted for the (subjective) typical amount of
        any food eaten. Therefore, completely meaningless.

        You are back engineering the load to the index.

        • So rude, so sure of yourself, and so easily shown to be wrong, all at the same time:

          • Michael, from your link: “…estimates how much the food will raise a person’s blood glucose level after eating it.”

            Sounds pretty subjective to me.

            Again from your link: “Glycemic load estimates the impact…” “…watermelon has a high GI, but a typical serving of watermelon does not contain much carbohydrate, so the glycemic load of eating it is low.”

            See that? “estimates” and “typical serving.” Pretty subjective, no?

            And how ’bout this whole paragraph, again, from your link:
            “For one serving of a food, a GL greater than 20 is considered high, a GL of 11-19 is considered medium, and a GL of 10 or less is considered low. Foods that have a low GL in a typical serving size almost always have a low GI. Foods with an intermediate or high GL in a typical serving size range from a very low to very high GI. (This sentence is ambiguous and imprecise about the relation between GL and GI, whereas the last sentence of paragraph 1 says they are directly proportional to each other. One of them is wrong. Paragraph 3 supports paragraph 1, but with a different formula (dividing by 100). Furthermore, this paragraph cites no authority for the range of GL for low, medium and high.)”

            Subjective as all get-out.

            Now, what were you trying to point out?

    • Hey, Bill! I think a better guideline than the glycemic index would be the insulin index:

  2. Dr Fung A question: Insulin signals cells to store fat and burn glucose (and other functions). I have been struggling with the question of the function of insulin when on a ketogenic, Atkins type diet, since glucose is already quite low. As you point out in your blog that protein (certain amino acids) stimulates the release of insulin. The insulin index of fish is 59 and meat 51(white bread 100). Since on the ketogenic diet many of these dieters consume more than 100 grams of protein a day, I wonder what’s the function of insulin–lowering glucose would result in hypoglycemia? Secondly, does it still promote fat storage?
    Note: I recommend on my website a low protein diet of 35-45 grams. I am not sure if that is important and don’t stress it. I believe this is what Anthony Colpo recommends.

    • Great question, Jerome. Check out this post for a potential answer. (Not the only one, perhaps, but at least filling in some small piece of the puzzle.) Insulin is anti-lipolytic, anti-catabolic. We do need *some* insulin to counterbalance glucagon, otherwise we’d all completely waste away via unchecked catabolism of both adipose tissue and muscle. This is what happens in type-1 diabetes.

    • It’s my understanding that protein, like carbs, is also turned into glucose when digested, though not as directly and problematically (in general) as carbs are. I suspect that the degree and manner in which protein is converted to glucose is dependent on a lot of factors, just as the insulin response from eating protein is not so well understood (as I think Dr. Fung has mentioned a few times). If this is right (and I’m no expert), then it seems appropriate that the insulin response in a healthy body from eating a specific protein meal would be commensurate with the amount of BG resulting from that protein consumption.

      As for your second question, Dr. Fung has said many times that it’s all about the insulin. So I would think that to the degree your protein meal produces an insulin response, it is influencing the fat storage/burn equation toward storage. It seems there are lots of variables and unknowns in the link from protein to insulin, but the link from insulin to fat storage is not so problematic.

  3. Thank you!!
    — for explaining so many metabolic issues with such clarity!
    — Elegant & Concise

  4. Dr. Fung,

    No doubt you have already planned your future posts.

    You and Minger both mention the issue of oxidized LDL as related to diet.

    I hope that sometime you can address the dietary and metabolic factors that are involved in that oxidation, as well as in raising serum Apo B or LDL particle numbers.

    In other words, add to the story about diet patterns and atherosclerosis, body fat, and diabetes.
    Perhaps I have missed or forgotten what you have already posted about these issues.

    Again, thank you for all your help.

  5. Thanks for another insightful article. Yet again you have helped me to understand a little more of the overwhelming complexity of nutrition and its effect on health. I’ve been struggling to understand how I did just as well with calorie restriction as with LCHF plus IF. I just found the latter much easier to stick to because it did away with hunger and cravings. I knew that what I was doing worked in practice but I couldn’t find a theory that explained it. LCHF/IF came closest but now CIH is an even better fit. Thanks for putting it into words I can follow. I realise that I still don’t fully understand all the theories and I’m sure they will continue to evolve faster than I can keep up with them, but I am following your blog with interest and have already pre-ordered you forthcoming book.

  6. Tim Heineman

    I find Peter Dobromylskyj’s posts on Denise Minger’s startling findings the most cogent. Try on October 10th and 14th, 2015.

  7. Awesome post and insights, Dr. Fung!

    FWIW, I’ve been trying to get people to see the difference between the CIH and lipophilia (the HOT as you call it) for a while. 🙂 Here’s a bit of what I wrote on this subject if you’re interested…

  8. Well, your HOT explains how the Pritikin, Ornish, etc. diets work (when they do). However, Denise mentions extreme low fat diets like the one having cooked white rice, sugar and orange juice. The underlying mechanism must be something different for those and Peter over at Hyperlipid might have the right one outlined.
    Considering all this, adherence and efficiency appears to be a key differentiator for those with serious insulin resistance. I believe more people find LCHF easier to follow and available data also points to higher efficiency, VLC working in ~90 % of people and VLF working in ~60 %.

    Dr. Jason Fung: I agree. I think Very Low Fat diets are difficult to follow if you also heavily restrict refined grains and sugars.

  9. Peter Lawton

    Jason, thanks for your recent calm commentaries on what Denise has invited us to think about, for pointing to what factors are at play. I was happy with the way Denise set about prodding us to think anew, but was quite surprised at the blunt and dismissive rejections put out by some well-known contributors to the “Paleo/ LCHF / Insulin /Real food” conversation. It was scary to see how quickly some apparently thoughtful, progressive people could jump overboard in their haste to write-off an invitation to think.

    • I could see how they could do that. For me, I used to keep my total fat content very low (less than 10% by calories). I basically ate very little meat, no dairy, and only (very) low fat products. It was a freaking horrible thing to do for my body. There may be people who can survive on such a diet, but for me, my blood sugar was out of control (I didn’t know it then, though). I would get angry, depressed, and was constantly hungry. That’s no way to live. It’s also one of the reasons I tried the Atkins diet, which immediately made things better. It took me years to convince myself that a high fat diet was good to eat, however.

      So, from my perspective, I couldn’t go back to eating a high carb, low fat diet, even if there are studies indicating these are “good” in some way. And since my personal experience is bad with these types of diets, it’s going to be hard to convince me otherwise.

  10. This explains why the weight I lost at 35 (eating cheaply with whole foods from all over the spectrum) was a whole lot easier to lose than the weight I lost post-menopausal (with LCHF and fasting)–INSULIN! I wasn’t nearly as resistant then as I am now.

    Dr. Jason Fung: Absolutely. This is the time-dependence of obesity, which everybody intuitively knows, but is not explained by either the Calorie paradigm or the Carbohydrate paradigm.

  11. How come glycemic load of white bread and whole wheat bread are equal if the added fibers are “protective”?

    At least I get a much longer period of satiety by eating whole-grain bread (which I don’t either any more in principle, but…) compared to white bread.

  12. Thanks for that shout-out on yet another fantastic post! Couldn’t agree more with what you’ve written here. I’ll be addressing similar issues in the final installment of my insulin series. It’s the *insulin,* but that doesn’t automatically imply it’s the carbs and *only* the carbs. People want black & white answers, but these things are multifactorial. Unfortunately, you can’t do multifactorial in 140-character sound bytes. It requires multiple blog posts and/or entire books, and, sadly, the people most likely to read those are people like yours and my readers, who *already* know how this stuff works and want to learn more, rather than the people still dutifully following the misguided advice of the ADA and similar organizations, who might benefit from it most.

  13. […] Sourced through from: […]

  14. I grew up on Okinawa and I can assure you that Okinawans are omnivores. Their diet is neither “starch based” nor primarily vegetarian–they eat rice, planet of vegetables, meat, fish, and seafood. The primary meat is pork and I can assure you they don’t discard anything–lard is a common cooking fat. Fish and seafood are abundant surrounding this tropical island. For a glimpse of traditional Okinawans foods (setting aside “taco rice” which is an Americanism) check out Anthony Bourdain’s “Parts Unknown” episode that recently aired on Okinawa.

    There are very poisonous snakes on Okinawa (habu) and Okinawans have even resorted to eating that meat when none else was available. Okinawa suffered some of the fiercest battles in WWII. The people who are elderly lived through that time, when many died. Food was scarce during the war (enforced fasting). Those who survived have great resilience and they are blessed with good air, good food, a relatively stress free lifestyle (now). They also drink a lot of tea and are moderately active every day. It is neither their “vegetable-based” diet, their “starched-based”, nor even their “pork and seafood-based” diet that is responsible for their longevity. It is the sum of many things.

    Dr. Jason Fung: Agreed. It is the sum of many things – just not the Taco Rice. LOL

    • Personally, I think stress has an incredible amount to do with it, particularly for heart disease.

  15. Dr. Alister Frayne MD

    Hi Jason,
    Agree totally. For those with the time, Dr. Robert Lustig’s book “Fat Chance” is excellent reading along the same lines (He’s a neuroendocrinologist from UCSF). The role of Ghrelin, Leptin, peptide YY, cortisol remain to be clarified, but can be summed up as “eat real food”.
    I often measure Insulin levels in my patients, and am constantly surprised by high Insulin levels, in the context of normal or marginally elevated blood sugars – simple dietary advice (steering these folks away from high “Insulinemic Index” foods) is true primary prevention, and they are able to reverse away from beta cell exhaustion (which is where they are ultimately headed), lose weight effortlessly, reverse fatty liver, lower BP, and generally feel much better.

  16. I like your expanded model, but I’m still not sure it can explain all the data. How about the people who eat lots of *refined* carbohydrate and stay very lean. The Kempner Diet may be one example (if your reduced calorie theory of Kempner doesn’t hold up). Another may be Asian diets quite high in refined white rice.

    I personally know many Asian people who consume impressive quantities of rice three times a day yet remain incredibly lean and seemingly free of metabolic disease over decades. Yes, they do eat some vinegar and veggie fiber, but I doubt that can fully explain their resiliency to all that refined carbohydrate. They do not eat much fat. They do not fast. They do not exercise.

    And they do not eat sugar. Perhaps that is a key, as Gary Taubes has speculated. But at least on face value, like Kempner they seem to be a paradox for the appealing theory that “the toxicity lies in the processing, not the macronutrient composition.”

    • “if your reduced calorie theory of Kempner doesn’t hold up.”

      I’ve looked at a number of sources, and at least for people who needed to lose weight, it does seem to hold up. It wasn’t an eat to appetite approach when weight loss was desired. Sometimes people are described as losing weight spontaneously on the diet–but calories were increased for these people (probably where the sugar comes in, it increases calories, but doesn’t increase the protein load that Kempner was so concerned about).

      Dr. Don Rosati and his wife Kitti carried the ball for the rice diet after Kempner. Here they’re being interviewed, Dr. Rosati gives a history of how the diet started being used for obesity. He does say that it was hard for some people to maintain bodyweight when the diet was originally used vs. malignant hypertension, patients weren’t generally overweight to begin with. Later in the 70’s when obesity became more common, he says Kempner adapted the diet for weight loss–800 to 1000 calories a day. Not much need to dig into theories involving nicotinic acid to explain how a plan like that might reduce insulin, very little food equals very little insulin.

      And there’s this personal anecdote, from when the Rice Diet program shut down;
      Under the rigid discipline of Kempner, enrolling in the Rice Diet residential facility was like entering boot camp for the overweight, said Jean Renfro Anspaugh.

      “One ate rice and fruit and walked. The staff didn’t care what you thought, only what you ate and how often you exercised,” said Anspaugh, whose book about her experiences with the program is called “Fat Like Us.”

      Anspaugh, 59, of Fairfax, Va., knew of Durham and the Rice Diet from celebrities discussing weight loss on television in the 1960s. That’s where she headed, dropping out of law school in Sacramento, Calif., within weeks after her excess weight collapsed a patio chair. She lost 70 pounds in four months on the Rice Diet, but ran out of money before reaching her target of another 50 pounds, and got a job with Kempner’s team.

      She believes the Rice Diet Program suffered from the multiplying number of diet books, the difficulty having insurance cover time at a diet retreat and the popularity of gastric bypass surgery.

      “Why go to Durham and spend all that money eating rice and fruit and starving to death when you can go to your local hospital and have it done and have your insurance pay for most of it?” Anspaugh said.
      One caveat to all of this is that the diabetic retinopathy studies were much earlier. The next is from a nitrogen balance study of the Rice Diet, done in 1949;
      This does not mean
      that the caloric supply is fixed at 2000 calories. It
      varies according to the varying indications for
      weight gain or weight loss in the individual patient.
      The nitrogen and total caloric intake can be
      controlled by altering the rice and sugar fractions
      of the diet.

      So clearly for any overweight diabetics, at the least, calories were restricted from early on, though we don’t know by how much, although we do know that it was very considerable in the 70’s and onward. But… does Kempner strike you as a man who was, at any time in his career, into half-measures?

    • There is probably at least some degree of genetic influence. There are plenty of groups that *don’t* do well on high amounts of refined grains.

  17. Of note is that Nathan Pritikin committed suicide after developing leukemia and its complications, so yes, he avoided CVD. But how did his diet affect his overal health? Or his mental health?
    In our n=2 experiment on the Pritikin diet my husband and I had opposite results. He lost 27 lbs on that diet, and I gained a couple. Why such different outcomes? Genetic and our baseline conditions come to mind as possible explanation. He grew up on an American diet, with more sugar and fat and processed foods. He still ate a typical American diet of more processed food, more sugar and more fat. The Pritikin diet was for him a significant improvement, and probably lower calorie compared to what he had been eating. While I grew up in Italy with less meat, almost no sugar and still ate a Mediterrean diet, dessert once a week made from scratch with real ingredients. In fact I was quite scornful of the typical American fare. But in addition to food snobbery, the most important difference is that we had insulin resistance from different causes. He from eating sugar and fake foods and oils, while I had undiagnosed sleep apnea with its attendant insulin resistance. Eventually I would gain a lot more weight. In any case, Same diet, similar metabolic situation but from different causes resulting in weight loss for one, weight gain for the other. Though this is strictly anectodal, it does point to the fact that to understand disease states, case variations are equally important to large scale, population based studies. In fact without taking into consideration those individual variations, the study design is incomplete and will give contradictory results. As in this analysis of two diametrically different diets.

  18. […] du dock Jason Fungs inlägg om det hela där han tar upp sin modell, som Kostdoktorn säger att han modifierat, så är det […]

  19. If you look up the foods which have a measured insulin index, your point is well made:

    Tortilla, white, corn (San Diego Tortilla Factory, Australia), insulin index 36, carbs 47

    Beef steak, grilled (Australia) : insulin index 37, carbs 0

    Snickers bar (Masterfoods, USA) : insulin index 37, carbs 29


  20. The problem with using the glycemic index is that we don’t all respond in the same way; blood sugar and insulin response varies from person to person. One person might be able to eat hoards of white rice and have a very small reaction, whereas another person might have a huge insulin response. I do agree that “it’s all about the insulin”, and I also agree that there’s no one right diet for everyone; however, many of us are metabolically damaged, and for us it absolutely IS “about the macronutrients”. This would not apply to people like the Hazda, who are healthily born into their whole foods diet. Industrial foods have been around for quite awhile now, so each subsequent generation is affected- (ever hear of “Pottenger’s cats”?

    Which brings me to my next point: I don’t believe this disproves the “Carbohydrate Hypothesis” at all. People like Gary Taubes and Phinney and Volek don’t say we all “have” to follow a low carbohydrate diet; what they do say is that people with carbohydrate intolerance (or people who are “metabolically damaged”) do better when they lower carbohydrates; (Gary Taubes even admits that low carbohydrate diets work for most people, but there are a subset of people who do better by limiting fat). Taubes was mostly against high levels of refined carbohydrates and sugar, which I think we would all agree with. Even Dr. Atkins believed that each person had a different level of carbohydrate tolerance, which was the main premise of the maintenance phase of his diet. What all of these people DID do was defend low carbohydrate diets, explaining that they are not “dangerous”, even for the general public, as many “authorities” have led us to believe. I think it’s some Paleo bloggers and followers who have maybe taken this to the extreme, misunderstanding what was actually written in the books. I’m not trying to “blame” anyone here (so I hope nobody takes this the wrong way)…’s a relatively complicated subject, so people do tend to simplify it.

  21. I think it would serve everyone well to go back and reread Protein Power. It has a good discussion of insulin and the effects of carbs, protein, fat, and their various combinations on insulin levels as well as glucagon levels.

  22. Surely the most important thing out of all this though is that whilst the Pritikin diet may well work (by reducing insulin) in similar ways to a LCHF diet, in terms of long term adherence a LCHF diet will nearly always win out. Simply because it achieves the insulin lowering effects while at the same time allowing people to eat more calories that satisfy hunger. Pritikin and similar diets the dieter will almost always end up feeling hungry and fall off the wagon in the longer term. LCHF and IF (and a combination thereof) allow one to adhere better because of the absence of hunger (in general). And it is the adherence to a long term lower insulin diet which will make one lose weight and keep it off. Right?

  23. Martin Willias

    A very recent study underlines Dr Fung’s hypothesis that sugar is much more harmful than other carbohydrates. All markers of metabolic syndrome were reduced in as little as nine days, and insulin by a third. Note that the benefit occurred even though simple sugars were replaced by bread, bagels, etc.!

  24. Thanks for the last two post Dr. Fung. The calories in, calories out theory is like a broken clock. It is right twice a day. The CICO theory is right only when the foods your are currently eating work well with your metabolism at that moment in time. But if your food or metabolism change, then metabolic disease will follow.

    However, the hormonal theory is right all of the time and thus encompasses the CICO.

  25. BTW, the CIH may be soooo 2010 for Dr. Fung. But for me it is soooo October, 2015.

    I would also give a shout out to Dr. Kraft via the excellent The Fat Emperor,

    My new campaign slogan is It’s The Insulin, Stupid. Thanks, Amy.

  26. Hi Dr. Fung,
    I enjoyed this post. I followed an Ornish/Fuhrman style vegetarian diet for many years before switching to LCHF. As someone alluded to earlier, the problem with these diets is you are always very hungry, because you can’t get enough calories from fruits and vegetables. If you want to be a vegetarian and you are afraid of fat you end up eating lots of bread, pasta and other grains to ease the hunger. I did this and peaked with an A1c of 7.4 before cutting back to 25g of carbs a day. The way I interpret glycemic load is that the 25 grams of carbs have to come from low GL foods, which effectively means salads, veggies, and a few berries.
    thanks, Jeff

  27. As a T2DM, I do better on a LCHF (Atkins-like) diet vs a high fiber, high carb, low fat (Pritikin-like) diet. My LCHF diet still has lots of fiber due to the judicious use of 1/2 cup of beans and an avocado or two per day. In addition, I eat 4 servings of green vegetables per day.


    The above link is interesting. The writer (you should read *all* his posts!) kicks around some ideas about the metabolic “magic” which may be going on under extreme low-fat. Be aware the first part of the post is very heavy with cellar biology and biochemistry. You can skip that to get to the end conclusions, which I think are a reasonable response to the study that Minger is looking at.

  29. Dr. Samir Anadkat

    Dr Fung
    I am your fan. And I am deeply greatful for the comprehensive knowledge you have imparted in my understanding of obesity.
    My question- I routinely do FBG and Fasting Insulin before prescribing dietary therapy for my patients. I also rule out thyroid (by TSH). But there are few patients where IR is quite low, there are no obvious signs of high cortisol and they are still obese.
    1. Do you think these kind of patients will benefit more from lower fat in diet vs lower carb? (Here high fiber, low animal protein, use of vinegar and fermented food, use of IF …. remaing same between two approach.
    2. Is there any non insulin dependent theory of obesity for small group of patients?
    3. What are better tests to screen for stress as driver for obesity apart from history and 24 hr urine?
    Thanks in advance.

Leave a Reply