Avocado Raises Blood Sugar

Tim Lundeen writes:

We [Tim and his partner, Alexandra] first noticed that eating avocados raised our blood glucose when we were on a low-protein/low-fat/high-fruit nutrition plan. After 1/4 avocado each, we would both have fasting glucose of 95-99 instead of 80-85, with the effect lasting for about 4 days. It was quite repeatable, so we stopped eating avocados. We speculated at the time that it was due to the omega-6 content of the avocado fat.

We just tried avocado again with more typical nutrition, with about 25% protein, 25% fat, 50% carbohydrate with very low fructose, thinking that because we were eating more fat the effect might not be so pronounced, but saw the same elevated fasting blood glucose as before.

After some more research, we found out it is because avocados contain a sugar called mannoheptulose, which causes temporary dysregulation of blood sugar.

Mannoheptulose was first isolated in 1917. Mannoheptulose has been proven to be present in many foods, but is found most abundantly in the avocado (La Forge 1916-1917). In 1957, the first research was published in the Archives of Biochemistry and Biophysics (Volume 69, page 592), suggesting that avocado extract blocked normal insulin secretion. In 1963, it was demonstrated that avocado extract blocks glucose-stimulated release of insulin (Nature, Volume 197, page 1264). By 1967, low doses of avocado extract were found to inhibit both pancreatic secretion and synthesis of insulin without eliciting measurable hyperglycemia (high blood sugar) (Nature, Volume 214, page 276). This finding was significant because it demonstrated that a controlled dose of avocado extract could suppress pancreatic production of insulin without inducing a diabetic state. [http://www.health-marketplace.com/p-Obesity-3.htm]
The problem with this is that your cells don’t absorb nutrition because insulin is reduced, so we have strong cravings for food, feel extra hungry all the time, and have been eating about 50% more calories to feel full. The net effect is not a good feeling…

This makes sense. And it is methodologically interesting. Spending zero research dollars, Tim and Alexandra learned something important about blood sugar control that the rest of the world seems not to know. (Except perhaps the researchers who did the avocado extract research.) None of the research articles they mention make clear the practical significance of the effect. To say that avocado extract does X doesn’t tell you how much avocado you need to get the effect.

When I google “avocado” and “blood sugar” (together), the first page of links all claim, at least at first glance, that avocado lowers blood sugar. But that’s just the internet. (Although Google is supposed to put the most reliable links at the top.) Then I went to the most authoritative possible source on what we should eat: the American Journal of Clinical Nutrition. I found only three articles that mention avocados in their title or abstract. None was about this effect. I also looked in Eat Drink and Be Healthy by Walter Willett and the Harvard School of Public Health. Nothing about this effect of avocado.

17 Replies to “Avocado Raises Blood Sugar”

  1. While the implications of mannoheptulose are interesting, the results from this n=2 trial and fairly worthless, seeing as how the couple utilized high carb diets both times.

    “The problem with this is that your cells don’t absorb nutrition because insulin is reduced, so we have strong cravings for food, feel extra hungry all the time, and have been eating about 50% more calories to feel full. The net effect is not a good feeling…”

    this statement reflects the results of their diet, and subsequent chronically elevated insulin, not the avocado, accurately. I regularly consume up to two whole avocados daily in conjunction with a high fat, moderate protein, low carb diet; fasting trigs are extremely low, and when I’ve tested my BGL, it has always been within normal range (5.0mmol/L). Satiety is extremely high with no strong cravings, ever. Perhaps Tim & co. need to analyze the whole picture, and not just a piece.

  2. So, if Gary Taubes is right that insulin is the primary regulator of fat storage, and if avocados lower insulin levels, then it makes sense to eat avocados if you’re trying to lose weight? (Or would doing so make you more hungry than normal, which might make you eat something that would then counteract the insulin-suppressing effect)?

    I’m trying to parse the practical significance here, in terms of weight loss.

  3. Interesting, from wikipedia:

    Roth and his team (from P&G Pet Care, Wayne State University, Southern Illinois University and the Pennington Biomedical Research Institute) have been mining avocados for an alternative — MH (for mannoheptulose). It’s a fairly simple sugar with a 7-carbon backbone.

    When fed to mice in fairly concentrated doses (roughly 300 milligrams per kilogram of an animal’s body weight), it improved insulin sensitivity and the clearance of glucose from the blood. Meaning it helped overcome diabetes-like impairments to blood-sugar control. MH supplementation also improved the ability of insulin, a hormone, to get cells throughout the body to do its bidding (and that’s a good thing).

    MH revved up the burning of fats in muscle. That’s the opposite of fat deposition and something that these scientists note “would be an expected effect of a calorie restriction mimetic.”

    Treated mice also lived longer — some 30 percent longer than untreated animals. And they were happier, I’m guessing, because they didn’t have to give up most of their chow to achieve this life extension. Indeed, their food intake and weight matched that of untreated mice.

    Roth, G., et al. 2009. Mannoheptulose Glycolytic Inhibitor and Novel Caloric Restriction Mimetic (Abst. 553.1). Experimental Biology 2009, New Orleans (April 19).

  4. i started eating avocado (1/4) in a salad and found that it satisfied my hunger to the extent that eat less over all; also, the effect seems to last a long time, i.e., i can postpone when i next eat. i haven’t measured by glucose, but i’ll try.
    I think this points out the problem of trying to extrapolate from the experience on one person. the body is complex and reactions to food etc.. will likely vary; it is unlikely that any underpowered “study” (in this case really an anecdote) is very meaningful.

  5. So, they ate an evolutionary excessive amount of carbs, and then found that down-regulating the hormone that absorbs carbs caused problems. That’s… not surprising.

    Learning that avocado down-regulates insulin is very useful though. i’ll be sure to eat it more often, as producing less insulin is very healthy as long as you aren’t consuming a ridiculous excess of sugars that need to be mopped up (I’d say anything in excess of 25% carbs is suboptimal, let alone twice that)

  6. Mike, perhaps your fasting blood sugar would be lower — closer to 84 — if you ate fewer avocados. Closer to 84 is better. I eat the same diet as you (high fat, med protein, low carb) and agree with you about its benefits. I agree that if you eat such a diet avocados should make less difference.

    Caleb, that’s a good point. But a large fraction of the world eats an evolutionary excessive amount of carbs. So this has practical value.

  7. If you eat avocados on a regular basis, your body will adjust to the reduced levels of insulin: your cells will increase the number of insulin receptors they express, and you’ll stop being hungry. When we noticed this effect, our elevated blood sugar went away after 4 days, which is exactly how long it takes your cells to rebuild these receptors and calibrate their needs to changes in their environment.

    However, it seems likely that reducing insulin levels this way is not healthy. If our bodies are designed to produce a certain level of insulin in response to a particular glucose/protein load, then reducing it is likely to have side effects even though the overt hunger and elevated blood glucose goes away in 4 days.

    With regard to the level of carbohydrates in the diet, the pancreas has sensors for both glucose and glutamine, so produces insulin in response to both carbohydrate and protein intake until both are back to baseline. If the effect of mannoheptulose was only on glucose sensing, then it is possible that you wouldn’t see a blood sugar increase resulting from combined protein/carb intake. My prediction is that mannoheptulose acts directly on insulin output, not just on pnacreatic glucose sensing, because we do see glucose increase with combined intake. Very hard to say what the balance is without doing the experiment and measuring insulin levels with all protein vs all carb intakes, but would be interesting to try it.

    Also, re the “correct” level of carbohydrates in paleolithic diets, there is a lot of debate and these numbers are constantly revised. There is a good paper in the European Journal of Clinical Nutrition by Eaton et al, “Paleolithic nutrition revisited: A twelve-year restrospective on its nature and implications”. They retrodict paleolithic macronutrient intake as 37% protein, 22% fat, 41% carbohydrate, which is consistent with studies of modern-day hunter-gatherers.

    Further, there are cultures today that eat very high carbs (the Okinowans traditionally ate 10% protein, 10% fat, 80% carbs) who have excellent blood sugar control and live long healthy lives (the Okinowans have one of the highest percentages of centennarians of any culture).

    Our own requirements for any nutrition plan we adopt is that we have fasting blood sugar of 83 and peak blood sugar after meals of no more than 120, ideally less. Our current diet does this, except for 4 days after we eat avocado 🙂

  8. I both want to criticize and defend Mr. Roberts here. The criticism is this; the sample size is too small to take it as Gospel. The results must be replicated.

    On the other hand, new hypothesis/avenues of research must come from somewhere. In regards to diet/exercise/sleep or any other endeavor where it is easy for people to play around with marginal changes and record data, amateur self-experimentation is likely to prove very fertile ground for new ideas. The proper response of a scientifically inclined skeptic would be to ask if the results are repeatable, not to discount them because the bearers don’t wear the priestly vestments of a white lab coat.

    Also, Seth, one of the values of a liberal education, if it is an education that includes an emphasis on the history of math, science, literature, and philosophy, is to free the mind from prejudices and dogma and to recognize dogma and prejudice when one sees it. A liberal education, unlike some other forms of education, attempts to inculcate the habit of using reason as the ultimate arbiteur of truth, as opposed to authority, mysticism, or emotions.

  9. Tim, thanks for your data rich and nuanced response.

    We both agree that there is a wide variety of macro nutrient consumption that can be healthy. It’s more the type of carbs and fats that are important at the proximal level.

    I’m a little skeptical of using modern hunter gathers to extrapolate the evolutionary environment, as those who have survived usually have done so because they live on marginal land that others didn’t feel was worth driving them off of, so the type of food such land supports may not be representative of the past.

    I do consider these diets great baselines, though once again, I’m not convinced they’re optimal. My personal philosophy is to only get enough carbs to fuel the brain and muscles with glycogen (i.e. avoiding ketosis most of the time), only get enough protein for body maintenance, and use fats to meet all other energy needs.

    The main reason is that protein and insulin restriction seem to be the key mechanism in all the longevity studies. Insulin seems to activate the pathways that result in aging, and most of the interventions that increase lifespan, such as curcumin, avocado, rapamycin, all seem to share the common factor of down regulating insulin pathways.

    I don’t agree with the second paragraph, as hunter gather diets vary through the seasons and food availability, and I believe we’re well evolved to handle such variations. Hormeosis suggests it’s even good for the body to have to respond to such intermittent stressors.

  10. i just got my glucose reading back; it’s 83 (taken at 3:00 pm, probably 3 hours after eating); my reading before this (when i wasn’t eating avocado every day) was 100 (it was fasting). i guess i’m going to continue eating avocado (altho i also take cinnamon in the a.m.).

  11. @ Aaron Blaisdell:

    and perhaps also, “what is your exercise regimen”? I wonder about this when I read that Seth eats 60 grams of butter a day.

  12. Nansen, I usually take a one-hour walk every day. Sometimes more. The butter greatly reduces my appetite, I find.

    Paul, I think all education tends to “free the mind from dogma” in the sense that the more you know, the more beliefs you have to choose between. I don’t understand how the history of math is relevant here. Sure, learning the history of science shows how cherished beliefs (such as the earth is the center of the cosmos) have been wrong. But science isn’t the focus of liberal education. I suspect MIT students know more about the history of science than Harvard students. Liberal education, in practice, means a large dose of humanities. How studying this or that novel or listening to an art-history professor’s obscurities tends to free one from dogma I fail to see. I think Veblen was right about higher education: Professors try to show off their uselessness. In all areas they try hard but in the humanities it’s easier.

  13. Is it possible for avocado to temporarily give you symptoms of diabetes? I have frequent hunger and urination regularly but it seems after a couple of days of eating avocado both are exacerbated, especially the urination. I went to the doctor and had a fasting blood glucose of 99. I am very fit and eat a super clean diet yet it appears I am knocking on the door of diabetes. I should also mention I have chronic insomnia and that might play a part.

Comments are closed.