Insulin Index – Hormonal Obesity XXIII

Insulin levels had always been assumed to go up or down with the blood glucose levels which led to the glycemic index which had always been assumed to be a surrogate measure of insulin levels.  The glycemic index did not turn out to be as successful in weight control because glucose does not drive obesity.  Insulin drives obesity.Insulin Index1 With the insulin index, it was realized that only 23% of the variability of insulin response depends on the glucose.  In other words, how much the glucose increases only accounts for 23% of the insulin response.  Even taking into account the other macronutrients fat and protein, this only accounted for another 10% of the insulin response.  The vast majority of the insulin response is still unknown. Some of the factors that are suspected or shown to affect the insulin secretion include presence of dietary fibre, an elevated amylose/amylopectin ratio of the starch, preserved botanical integrity (whole foods), presence of organic acids (fermentation), addition of vinegar (acetic acid), and addition of chili peppers (capsaicin).  We will explore some of these factors in future posts. Nevertheless, the main point here is that there are many factors in the co-ingestion of foods that affect insulin.  Things are about to become very complicated.  The simplistic “Carbs make you fat!”, or “Calories make you fat!” or “Red meat makes you fat!” or “Sugar makes you fat!” sort of arguments simply are not able to capture the complexity of the human condition of weight gain. Among breakfast cereals, there is wide variation in the insulin response.  All-Bran, with its high fibre and promise of colonic regularity, seems be stimulate insulin much less than cornflakes for instance.  Protein containing foods turn out to be surprisingly potent at stimulating insulin.  Beef and fish, for instance, have virtually no effect on the blood glucose levels.  Yet they stimulate insulin almost as much as most cereals.Ins Index Carb Nevertheless, we are able to discern some general patterns here.  If we look at carbohydrates, there is a noticeable trend here.  As the total grams of carbs goes up, the insulin level also tends to go up. This is what we expect.  This is the basis of many low carbohydrate diets from the original Banting diet of the 1850s to the modern day Atkins diet and its many imitators. It has been recognized for well over a century that the amount of carbohydrates consumed plays a role in the development of obesity.  However, this study indicates that the correlation is far from perfect.  The correlation between carbohydrate content and insulin is relatively low (23%).Ins Index fat Dietary fat, on the other hand tends to have a lowering effect on insulin scores.  As the amount of fat increases, the insulin score tends to decrease.  Fat also tends to have minimal effect on blood glucose, so this is pretty much as we expected as well. Carbohydrates tend to raise glucose and insulin.  Dietary fat tends not to raise glucose and tends not to raise insulin as well.  So far, so good.  The surprise here is dietary protein.   Looking at the protein, there seems to be a slight trend here as drawn by the line.  As protein goes up, the insulin score tends to go down.  But you can see that this relationship is mostly driven by 2 main outliers at the extremely high protein intake levels (40 and 60 grams).Ins Index Protein If we remove these outliers, then it seems that there really is not much relationship between the amount of protein and the rise in insulin.  Some proteins raise insulin a lot and some proteins raise it only a little.  Dairy protein especially whey, seems to be the biggest culprit here. The implication is that some protein containing foods can lead to obesity while other protein containing foods are not likely to do so.  Here the data become extremely murkily and there are numerous inconsistencies in the literature. In the study “The acute effects of four protein meals on insulin, glucose, appetite and energy intake in lean men” published in the British Journal of Nutrition in 2010, insulin levels were measure in 30 men after a test meal.  Four different proteins were tested – eggs, turkey, fish and whey protein.Protein Insulin 1 Whey protein is one of the proteins found in dairy.  Cow’s milk protein is composed of 80% caseins and only 20% whey.  Casein is mostly responsible for milk allergies.  Whey is well known as the byproduct left over in cheese making.  Milk is separated into curds and whey.  Whey protein is often used by bodybuilders as it is high in Branched Chain Amino Acids (BCAA) that are felt to be important in muscle formation. Compared to the other proteins, whey resulted in the highest insulin levels.  This is despite the fact that none of the proteins had a very large effect on blood glucose.Prot Insulin 3 In this study, after the test meal, participants were given a buffet lunch 4 hours afterwards so that effects of protein on satiety could be measured.  The whey protein suppressed appetite to the largest degree.  It is certainly possible that whey protein may have two different effects on weight gain.  The increase in insulin may promote weight gain, but the decrease in appetite may suppress it.  Again, I will stress that I do not have all the answers here.  The field of study is new and ever evolving at this point. The propensity for dairy to cause insulin to increase was confirmed in the paper “Glycemia and insulinemia in healthy subjects after lactose equivalent meals of milk and other food proteins” published in 2004 in the AJCN. Here the effects of milk and particularly whey are seen clearly.  In fact, the increase in insulin far exceeded even that of whole wheat bread.  The effect on glucose is negligible but the incretin effects of the stomach (release of Glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide 1 (GLP-1)) are felt to be responsible for the increase in insulin.  We discussed the incretin effect in a previous post. The insulin stimulating effect of whey may also be seen when taken with foods.  A recent study in Israel gave whey protein to type 2 diabetics after a breakfast of white bread and jelly – a breakfast designed to increase sugars significantly.  Whey protein supplementation reduced blood sugar levels by 28%.  The incretin hormones total and intact GLP-1 increased 141% and 298% suggesting that the increased insulin is largely the incretin effect. Here’s the problem, though.  Blood glucose does not drive weight gain and diabetes.  Increased insulin does.  What happens to that blood sugar?  It is not removed from the body?  It is merely shuttled to the liver where it is turned into fat.  So, in effect we have removed sugar from the blood where we can see it, and put it in the liver, where we cannot.  Is that good?  Hardly The high insulin secretion may lead to the development of insulin resistance.  This can explain the findings of this paper “High intakes of milk, but not meat, increase s-insulin and insulin resistance in 8-year-old boys” published in the European Journal of Clinical Nutrition in 2005. Prot insulin 2 Here they studied eight year old children.  Given a test meal of meat or dairy, the insulin levels were measured.  Not surprisingly, the insulin response was higher to the milk.  After seven days of milk versus meat, the milk group started to develop insulin resistance.  This is, of course, the first step towards the development of type 2 diabetes and a key player in the time dependent effects of obesity. The fact that this develops after only 7 days of milk overfeeding is downright scary.  This data would seem to indicate that dairy consumption may lead to weight gain.  Is this actually true?  The truth, as usual, is far more nuanced.

Continue here for Hormonal Obesity XXIV – Is Dairy Fattening?

Start here with Calories Part I – How Do We Gain Weight?

See the entire lecture – The Aetiology of Obesity 3/6 – Trial by Diet

35 Responses

  1. Interesting research that only 23% of insulin response is due to glucose levels. Thanks for sharing. Some research is associating intestinal permeability with increased insulin response. What are your thoughts on intestinal permeability and insulin levels? Also the variable insulin response to milk and whey protein is fascinating. All foods should be tested for their insulin response and an “Insulin Index” created to replace the glycemic index. When it comes to the insulin response to dairy products, I would expect the insulin response to vary greatly with the dairy products depending upon fat content, protein content, possibly whey/casein content, as well as whether the food is fermented (kefir, real yogurt).

    • Great insight. The fermentation process turns out to be very protective. The lactic acid produced, similar to the acetic acid in vinegar may be protective by reducing gastric motility. Stay tuned.

      • Dr. Fung,
        Are you familiar with the GAPS diet? It specifically deals with healing the gut and digestive system however I wonder if it’s beginning stage that has meats boiled in broth with approved vegetables produces the incretin response. My weight loss has slowed down/plateaued since beginning the GAPS protocol.

    • Roger Bird

      Robin, my march toward diabetic health has been slowed if not stalled, and I think that it is because I drink so much kefir. I just finished off all of my kefir today as my one meal per day; it is actually more kefir than I usually drink, and I will see if my feet get fat and my open sore acts naughty. I know that kefir is loaded with whey and I think that it may still have the casein in it. I saw somewhere on the Internet that yogurt boosted insulin, so I see no reason why kefir wouldn’t. And I say this as a very strong supporter of kefir.

      Heck, it is standard operating procedure for fermenting kefir will separate into curds and whey, and the whey boosts insulin the most, I believe. This means that later when I reduce insulin resistance to zero that I can make and drink kefir and just allow the kefir to separate, eat the curds, which are really delicious, and compost the whey. (:->)

  2. I’ve heard that skim milk consumption does not help people lose weight. This article really drives home why that might be.
    If milk proteins raise insulin levels, but fats generally suppress them, then skim milk is about the worst possible way you can consume milk.
    Do any studies compare the insulin response to various levels of milkfat? i.e. Whole vs. 2%, 1% and/or skim?

  3. Insulin gets all the spotlight, but not many folk talk about the insulin: glucagon ratio which gives insulin levels context.

    High carbs and high protein meals both increase insulin, but the protein stimulates glucagon also whilst carbs suppress it.

    Of course that’s abstracted to macronutrients and we need to delve further into types of carb/protein, but point is it’s an important consideration.

    • Most likely you are correct. However, the amount of research into glucagon or insulin/glucagon ratio is many many times less that insulin so the answers are not known

    • Protein elicits a glucagon response because the protein-induced insulin spike causes blood glucose to drop due to glucose being taken up by tissues at the signaling of insulin. Glucagon signals the liver to release glucose to keep blood glucose steady.

      In the Gary Taubes Carbohydrate-Insulin-Hypothesis (CIH), protein should also be making people fat because it causes an insulin response and glucose uptake. Lean people such as body builders use protein (BCAA is an example) and carbs (even sugary carbs) to spike the insulin for the purpose of building muscle. Many ordinary people live off such a diet and are normal weight. For many people, they are not fattening carbs and fattening protein. The results are the opposite of that predicted by CIH. The real world is more complicated than simple theories such as CIH.

  4. I’m just wondering if an Insulin Index would be any more helpful than the glycemic index turned out to be. After reading this post, it sounds to me like the complexity of digestive process would make any index not very useful. My thought was how beautifully fasting bypasses all of that. And thank goodness too, because I don’t want to wait around another 20 years for all of it to get figured out. I was also very happy to hear that you will be talking more about fermented foods. The research and self experimentation that I’ve done has convinced me that for many people, a return to health is going to have to involve restoring and supporting the microbiome with homemade cultured foods.

    • I think you are absolutely right. There are probably much more to the story than we currently know. In 10 years, the story may have changed again, but again, fasting does bypass everything.

    • Dunno, I’ve got stuff here from decades ago demonstrating simple carbohydrate restriction is as good as extended fasting for much of the maladies we try to figure out. It seems we just keep forgetting or choosing to ignore the past. Eg:

      “Changes in plasma glucose, free fatty acids, ketone bodies, insulin, and epinephrine concentrations during fasting were the same in both the control [fasting 84 hours] and lipid [infusion fed] studies.”

      “The hormonal pattern switched towards a catabolic mode with a fall in insulin levels .. and a rise in glucagon concentration. A significant fall in triiodothyronine and rise in reverse triiodothyronine were observed, while thyroxine levels remained unchanged. The average levels of the most important gluconeogenic amino acids .. were reduced .. while those of the branched chain amino acids increased by more than 50%. Since these changes reproduce those observed after a few days of total fasting, we suggest that it is the carbohydrate restriction itself which is responsible for the metabolic and hormonal adaptations of brief fasting.”

      • The problem I’ve found though is that carb restriction isn’t simple. Not for me at least. Low carb absolutely does work because I’ve proved it on myself, but for me sustaining it for the long term was difficult. The advantage of the fasting is that it provides a little more leeway with the macronutrients. I would love to have made a bigger change in my eating habits by now, but after 53 years they are fairly ingrained. I am making progress, but it’s a relief to be able to have fasting to help me heal now while I continue to improve my diet.

  5. Paula Gaubert

    What about goat milk ? Does it get a free lunch ?

    • Roger Bird

      I doubt it. Unless goat milk is somehow devoid of casein and whey.

  6. […] we saw in our previous posts, protein containing foods may stimulate insulin partially due to the incretin effect.  These are […]

  7. Hi Dr. Fung, I’ve just discovered your site after reading about it on a low carb message board. I feel like I’m reading a cliffhanger novel! I love your writing style, and I want to get to the last chapter! I’m a post-menopausal woman who began to gain serious weight after suffering a bad knee injury at age 30. Within a year of the injury I had gained 50 pounds and it just went up and up and up from there until eventually I reached (highest recorded weight at doctor’s office) 375 pounds. I developed T2 diabetes, high blood pressure – everything seemed helpless.

    I went on a low carb diet back in 2006 and have continued it ever since. Low carb is so easy to stick to for a lifetime, ,but it hasn’t gotten me where I want to go. By 2009 I had lost 100 pounds, and have kept it off ever since. But in five years I have unable to shift any more pounds despite having at least 100+ to go! My blood sugar and blood pressure are still not where I want them.

    Reading your posts, and about the insulin index, make me wonder if there really is any hope for someone like me. Currently I’m in the wilds of rural maritime Canada on vacation, do don’t have an internet connection that supports watching any of your videos – so I only have the blog.

    I like the idea of the 24-hour fasts. I just wish I had a better idea of what I should eat during the eating periods! I’ve been trying your plan as best I can for 1 1/2 weeks so far (and 4 fast days). So far no improvements in blood glucose – though who knows about insulin, and I do feel better. I just want the next installments of the novel to come out! 🙂

  8. […] Continue here to Hormonal Obesity XXIII – Insulin Index […]

  9. It seems you are hitting me as a diabetic “high and low” (to use a boxing metaphor). As I was reading this page, I was musing about a study done in Israel regarding whey protein, then there you go and reference this same study. What I got from the study, is that whey protein could be effective in helping prevent post-prandial blood sugar spikes, so I go to the health food store and get whey protein for morning protein drinks. Now you tell me, from the same study, that whey protein is effectively spiking insulin making weight gain probable. Is that the correct take? For many of us laymen who are fighting diabetes, this is the problem of prospective–same study going two different directions. Correct me if I have misread this.

    • Whey protein increases insulin. This reduces blood sugar. However it also causes weight gain (insulin causes weight gain). I think this is a bad thing. I think insulin eventually leads to increasing insulin resistance, they key problem in type 2 diabetes.

      • Only when insulin is chronically elevated longer than it is kept low does it cause weight gain. When you have been eating high carb, high fat for a long time while living a low-activity life, you fill up your liver and muscle cells (your largest, but still LIMITED depot for carbohydrate in the body), and then the body must convert excess carbs into adipose fat. All that dietary fat that was eaten along with the high carbs was also stored in your adipose tissue due to the high insulin response of crappy processed and refined carbs. The only time you give your body a chance to burn any of that fat you just made and stored is when you are sleeping (basically), and you burn the fewest calories during this time. It’s simply not enough to overcome the amount of fat you are CREATING and STORING, so… you get fat (and insulin resistant due to your cells being filled with too much fat). You can reverse this optimally two ways — a highly clean whole foods diet that’s either low-carb, high-fat or the exact opposite ala 80/10/10. Either way works over time, but you have to ELIMINATE all the man-made processed and refined JUNK that cause high insulin spikes without a concomitant rise in glucagon to counter balance it. Fruits and veggies are filled with fiber and as such are released slowly into your bloodstream keeping overly high insulin spikes from occurring (fruit juice however is a no-no — all the fiber is gone). Meats and other proteinaceous foods can cause an insulin spike but with a somewhat opposite effect by releasing glucagon at the same time thus blunting the overall insulin effect. That is why it has been shown time and again that replacing carb calories with protein calories (up to a point) speeds up weight loss. Replacing carbs with fats also works (again to a point) because insulin is not triggered by eating fat and this replacement actually helps your body burn the fat you ate AND the fat you have stored over the years by keeping the all-important insulin to glucagon ratio LOW. The overall point is that OVEReating bad, insulin-spiking carbs (low-fiber, man-made, refined grains, sugars, etc.) and too much dietary fat at the same time is the primary reason people gain fat stores over time. Keep that insulin to glucagon ratio low and you will lose weight easily.

  10. ———“”The simplistic “Carbs make you fat!”, or “Calories make you fat!” or “Red meat makes you fat!” or “Sugar makes you fat!” sort of arguments simply are not able to capture the complexity of the human condition of weight gain. “”

    ———-“”However it also causes weight gain ‘(insulin causes weight gain)’. I think this is a bad thing.””

    Saying that insulin causes weight gain is a simplistic approach, right?….. These seem to be hypocritical statements…
    Please explain.

  11. Thanks Jason for highlighting the importance of the insulin index. Fascinating topic. People might find this post helpful to understand the effects of what we eat on the body’s insulin response. It’s received nearly 100,000 views in six months so it seems there’s a lot of interest out there in the topic!

  12. […] פוסט זה מוגש כשירות לציבור. הזכויות על התוכן שייכות לכותב של הפוסט המקורי. את הפוסט המקורי ניתן למצוא בכתובת הזאת. […]

  13. Oh my goodness! Great article, thanks! Seems Atkins was onto something with the ol’ egg fast after all!

  14. A phd girl wrote a thesis on insulin index. Download it here. It has a Food Insulin Index (FII) table of 146 foods. Bravo!

    Thesis is in two parts. Second part has table and pictures of food.

    • Kirstie Bell only wrote Food Insulin Index (FII is for 1000 kj), but only a little FID ( Food Insulin Demand), anybody got the FID list?

      FII is like Glycemic Index (GI), and FID is like Glycemic Load (GL).

  15. […] Ruoat, jotka nostavat insuliinia, eivät rajoitu pelkästään sokereihin ja muihin hiilihydraatteihin. Myös monet proteiinit aiheuttavat sitä. Paras esimerkki lienee maito, jota Suomessa käytetään eniten maailmassa. Vaikka se ei juurikaan nosta veren sokeria niin insuliinia se nostaa jopa enemmän kuin vaalea leipä. Insuliiniresistenssin maito saa aikaiseksi tehokkaammin kuin punainen liha (kuva lainattu täältä): […]

  16. Pablo Sanchez

    Dr Fung,
    Thank you for your interesting work.
    One question: if fat not release insulin, and carb release insulin unless protein, then the best diet, to health and weight, would not be a diet whitout protein, but moderate in carbs and fat? Like vegan mediterranean diet? Or only carbs, in case that the fat, a long term, increase the insuline

  17. Are the food Insulin Index scores the same as % insulinogenic scores? Thank you.

  18. is taking the BCAA when fast will do harm on the insulin level?

  19. […] mutta ainakin insuliiniresistenssin aiheuttajana se jää toiseksi maidolle (ks. tämän jutun lopussa oleva grafiikka ja linkki […]

  20. JustMy2Cents

    Looking at this is a mess.
    Pick a side for the diagrams, sheesh.

  21. Dr. Fung, thank you for your work.

    Your insulin index graph appears to have a serious error. The “beans” at bottom with the ultra-high insulin index appears to actually be jelly beans, not beans, which you would expect to have a low insulin index. See the insulin index graph here:

    which has “jelly beans” at about that level, and has lentils at an expected low level.

    • Richard H. Serlin

      Sorry, it appears that actually is beans, and not jelly beans. See:

      Beans, surprisingly, do have a high insulin index; I guess a lot of carbs. However, insulin index is insulin response per calorie, and beans are extremely filling per calorie, with all of the fiber, so it’s hard to get a lot of calories from beans, and thus a large total insulin response. Clearly, it would be nice to have a measure for insulin analogous to glucose load.

      And, it’s important to keep in mind that beans have important nutrients, and have been shown to be quite anti-cancer. Joel Fuhrman recommends them highly.

Leave a Reply