Remembering Kevin Moore

Registered Nurse/Certified Diabetes Educator Questions Gary Taubes About 'The Pizza Effect'

The people who read the “Livin’ La Vida Low-Carb” blog are some of the most educated, informed, and thought-provoking people in the world of health. From laypeople to doctors, dietitians to nurses, the level of expertise and experience on everything related to diet, health, fitness and nutrition is as vast and deep as the ocean. As a simple layperson who has educated himself about these things since 2004, I am constantly amazed by the insightfulness of the people who read my blog, listen to my podcast and engage in the debate of ideas about what healthy living is. Every day is a new classroom experience for me as new information is made public and greater learning and understanding commences. Many times this happens whenever I receive an e-mail like one that was recently submitted to me by a registered nurse/certified diabetes educator about something she read and heard from Gary Taubes.

The bestselling author of the game-changing health blockbuster in 2007 called Good Calories Bad Calories recently came back on my podcast for an interview about his upcoming December 2010 release of Why We Get Fat And What To Do About It. This was what precipitated the feedback from the RN/CDE regarding her concerns about a phenomenon known as “the pizza effect” where blood sugar levels soar abnormally after eating certain foods. She was concerned that Taubes was erroneously putting the onus of high blood sugar and insulin levels on carbohydrate consumption when dietary fat could also be playing a role in this happening. Here’s what she wrote:

Hi Jimmy:

Thanks for posting this interview.

I hoovered up Good Calories Bad Calories in 2007, but the past two years I have been leaning more towards page 397 and seeing people with diabetes improve blood sugars and lose weight on very low-fat diets without “the cost of increasing fat accumulation and obesity.” [Line 6 page 397]. So I wanted to hear what Gary has to say about this three years later.

Gary does briefly discuss basal insulin levels and macronutrients in response to a question you asked him when he stated: “The key fact here is, it’s not just the short term insulin secretion, it’s the long term level of insulin in your blood and the relative sensitivity and resistance of the different tissues to insulin.” Agreed.

However, I do find that dietary fats increase late post prandial and basal insulin needs and this insulin resistance can be seen especially in people with Type 1 diabetes who monitor their blood glucose closely. People with Type 1 diabetes are unable to compensate for increased insulin resistance by upping their insulin production, so when they become more insulin resistant, their blood sugar levels run higher.

Hours after eating a high-fat meal, people with Type 1 diabetes often find their BGs remain higher than expected. For many years, this has been unofficially termed “the pizza effect” and this hyperglycemia often extends through the night and shows up as an elevated fasting BG the next day. Here is just one example from the Internet: “Seeing a graphical depiction of blood sugar after a high fat meal was shocking. We could see the blood sugar start to spike hours after we expected.”

So when determining post prandial insulin needs for people taking insulin, it’s not just carbs anymore, proteins and fats as well need to be accounted for to keep blood glucose in optimal range. See the presentations on this made at this year’s American Diabetes Association Scientific Sessions here and here.

This review of recently published research illuminates a fascinating immune/metabolic connection between dietary fats, inflammation and insulin resistance. This may be a major contributor to “the pizza effect” and may also explain why some people are, as Gary says in the interview, “beyond the point of no return” and do not lose a significant amount of weight when they follow a very low-carb/high-fat diet. High levels of inflammation may be keeping these people insulin resistant and this prevents them from losing weight.

This e-mail from the RN/CDE reminded me of the controversy stirred up by Dr. William Davis at “The Heart Scan Blog” earlier this year when he asserted that consuming fat like butter will raise insulin levels and make you fat (and he recently rekindled the debate again with his post last week entitled “Butter: Just because it’s low-carb doesn’t mean it’s good”). As a diabetes educator, my reader is greatly concerned that people like Gary Taubes may be overlooking an element in the management of blood sugar and insulin levels that is not directly tied to carbohydrate. When I forwarded this inquiry to him for a response, I had a sneaky suspicion he’d find the culprit is still the carbohydrates. Here was his response:

Interesting comments, but “the pizza effect” URL she provides is, indeed, pizza, which they describe as a high-fat food, when actually it’s both high-fat and high-carb. And much of the fat in pizza comes from soft cheese which may elicit a significant insulin response. So it would be interesting to know what happened to this man’s blood sugar had he eaten the cheese (and pepperoni, ideally) but not the crust.

As for the review article she linked to, there’s a similar problem. Carbohydrates are converted in the liver into saturated fatty acids. Dr. Jeff Volek from The University of Connecticut has published a lot of studies on how saturated fat levels in the blood are determined far more by the carbohydrate content of the diet than the fat. So is it the carbs? Could be.

Not that I’m saying she’s wrong, but the evidence she provides is based on the conventional wisdom (the review article talks about over-feeding and high-fat diets causing obesity) and then they draw their conclusions from that conventional wisdom, rather than understanding, for instance, where those serum saturated fatty acids might be coming from and what might actually be causing obesity.

Another thing she’s ignoring are the numerous clinical trials of Atkins studies, which are the best clinical evidence we have for what happens when people eat high-fat diets–they lose weight and they become insulin sensitive.

So, is it the fats or the carbohydrates that are leading to “the pizza effect” phenomenon where insulin spikes and stays elevated for hours on end? From what we know from the various studies that have been released in recent years, carbohydrates most definitely do this at varying levels depending on the fiber content and what they are consumed with. However, with dietary fat, under what circumstances would they have any measurable impact on blood sugar or insulin levels in the absence of significant amounts of carbohydrate or protein? I’m not seeing it happen but I welcome the input of anyone who has a theory they’d like to share.

For me, if you cut the carbohydrate and focus on consuming mostly fat with some moderate amounts of protein, then you are putting your body in a position to keep insulin levels low and, in turn, keep blood glucose levels stable. This will help create the right set of circumstances for your body to experience weight loss and reduced inflammation levels for optimal health. Avoid the pizza and other culprit foods that would lead to extended elevations in insulin levels and enjoy everything that livin’ la vida low-carb has to offer you. It’s as simple as that.

  • Guru

    Ok i have a question, has anyone heard of the insulin index study? and the fact that some high-fat, low-carb foods ellicited a disproportional insulin response to their carb load?

  • It seems when I read these studies, there is always a carbohydrate portion to the high fat diet. What gets missed is that you will get increase conversion of dietary fat in the presence of carbohydrate due to a high insulin release.

  • Jonathan

    Drinking a large amount of water can also cause an insulin release. It’s not that insulin is released at all, but for how long is the insulin response happening due to continual digestion of carbs over a time period. If the carb load is near zero, then the insulin release would be short lived or otherwise the person’s blood sugar would drop too low.

  • Elenor

    I eat pizza — without the crust: I pull the toppings off (usually just extra cheese, green pepper and pepperoni). I’m pre-diabetic (or maybe no longer, as low carb seems to be helping my blood glucose readings), and my blood sugar responds pretty normally to pizza-sans-crust. Just n=1, and not in a diabetic, but I’m always dismayed by people calling things “high fat” — that are actually high CARB and high fat… {sigh}

  • You’ve interviewed plenty of folks this year who seem somewhat skeptical of insulin as the main culprit. Stephan (whom I’ve a lot of respect for) among others, seemed more inclined to leptin. But until someone comes along with a tome as dense and well-researched as Taubes’ I’m going to stick with with insulin as the bad guy.

    I just took a look at the first study she pointed to (http://professional.diabetes.org/Abstracts_Display.aspx?TYP=1&CID=80834) It had 23 subjects, took place in Poland and wasn’t even spell-checked. OK, I live in Eastern Europe so I’m not trashing the Poles, or claiming to be an expert on languages, but I’m somewhat skeptical of the rigorousness of a tiny study that can’t spell “influances” correctly. And what was the point of the study? “We test the hypothesis that the fat-protein meal rises postprandial glycemia thereby should be considered in prandial insulin dose dditionally.” And how did they do this? “Patients in EG received dual-wave boluses for dinner -pizza (45 grams of carbohydrate and 400 kcal from fat-protein products it means 4 Fat-Protein Units-FPU)” The subjects apparently had low-carb pizza as a way to test their post-prandial response to fats and proteins. Am I the only one shaking my head in wonder?

  • Hey Jimmy,

    Have you ever watch those shows on Discovery Health or TLC such as Half Ton Man or Half Ton Dad? I notice those morbidly obese people eat a lot of white bread. Also on those shows those people eat a lot of pizza. Pizza is just mainly dough.

    I do not see a lot of people getting obese on eating meat but I see a whole lot of people get fat on eat pizza.

  • Anne

    Dr. Bernstein had a number of his diabetic patients drink olive oil without carbs and he found no blood sugar increase. He says fat has no effect on blood sugar level other than to slow down the effect of the carbohydrates. Perhaps that is the explanation of the “pizza effect”

    Fat is eaten with carbs and then the fat gets blamed for the blood sugar spike. Doesn’t make sense to me.

  • Peter Silverman

    Denise Minger alluded to an interesting question in your interview with her. If a diet that empasizes whole. unprocessed foods seems to have a health-giving effect, how can you tell if it’s because of the empasis on whole foods or the ratio of fats and carbs? She was talking about Colin Campbell’s diet, but the question applies just as well to a whole foods low carb diet.

  • Paul

    Am I missing something here? Is pizza a ‘high fat food’?

    In my experience, pizza is a method for restaurants and supermarkets to sell ‘bread’ at high mark up. The profit lies in maximising the bread proportion and minimising the toppings.

    The health agencies have a habit of describing anything with more than 5% fat as a ‘high fat food’. And then they draw the erroneous conclusion that the, say, 10% fat is the problem.

    As to the ‘pizza effect’, the fat probably gets mixed up with the carb in the gut and leads to a longer period of insulin spike, although with a lower peak.

    But surely, it would be quite easy to determine if it is the fat, or the carbs, by asking a patient to eat just the toppings, or just the crust, and then measuring the result.

  • Guru


    in the study, insulin levels were measured for 2 hrs after eating. the insulin score is based on the area under graph, which would take time into account i suppose? the area under graph would not be sufficient to illicit a high score if the insulin spike was short lived.

    perhaps the rise in insulin was due to gluconeogenesis during the 2nd hour? i dont think theres another way around it, specially that the subjects were fasting for 10-12 hrs.

    it would be nice if the insulin index is given more attention, further knowledge is important for low-carb diets, IMO.

  • Olivia

    Most of the time when diabetics are buckling down to Get-Healthy! they are dumping sugar and a lot of junk food; reaping lots of benefits from subtracting poison, then pointing at the low-fat aspect rather than the low-sugar.
    So why do people still point to fat out of all the other macronutrients involved? What about testing the effects when fat-only is consumed? Carbohydrate only consumed? protein only consumed? Then various mixes between the three – including with fiber and without. It’s all so speculative.

    Pizza makes for a horrible example of testing for specific causes. Meat lover’s or vegetarian? Is it one ingredient or another? Could an additive be compounding effects? How much protein has been consumed? How much fat, grams of protein, fiber, and digestible carbohydrate? How efficient is the subject at converting protein to glucose? Was the subject in ketosis previously? And on and on and on.

  • Where blood sugar control is concerned, there are two important considerations; the appropriateness of macro-nutrient intake and the level of Arachidonic Acid (AA) intake.

    For some diabetics, a high fat diet is not appropriate because they are not biochemically equipped to process large amounts of fat. Here’s an excerpt from http://www.dadamo.com/knowbase/PATHbase/depict.cgi?23:

    Intestinal alkaline phosphatase activity rises following the ingestion of a fat-containing meal, especially if the triglycerides in the meal are long-chain fatty acids. In a study of volunteers given different test meals, the after-meal rise in serum intestinal alkaline phosphatase activity was significantly greater following the long-chain fatty acid meal than following the medium-chain fatty acid meal, and significantly higher in Type O and Type B over Type A and Type AB. Paradoxically, it appears that intestinal alkaline phosphatase gives Type O and Type B metabolic advantages when they eat high protein meals. Studies show that the consumption of protein further increases the levels of alkaline phosphatase in the intestines of Type Os and Type Bs. Without protein in their diets, Type Os and Type Bs do not gain the benefits of the specialized fat busting enzymes in their intestines. This explains why these blood types can lower their cholesterol by adopting high protein diets.

    Nobody,of course, is equipped to deal with large amounts of AA. Interestingly, the article about dietary fats and insulin resistance supplied by registered nurse/certified diabetes educator did not even mention the role of AA in inflammation and insulin resistance. Here’s research on that: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2730166/ and http://www.jbc.org/content/269/1/639.short

  • As is so often the case, Mark Sisson (of Mark’s Daily Apple) has a high-quality post on the topic of the insulin index:

    The net takeaway is the preliminary findings were intruiguing and counter-intuitive (for those of us adhering to high-fat/low-carb diets). It is also true the original researchers appear to have abandoned the work. There have been no further studies since the premilinary results they obtained as students in the late 1990’s.

    It’s hard to know what to make of this silence, but the net for me is that it’s interesting but not terribly useful.

    Will certainly be on the lookout for more data, though.

  • This reminds me of criticism of fast food: sugary drink, starchy fries, starchy bun, sugary ketchup, and a little meat (sometimes cut with soybeans) equals a high-fat meal. Pizza: starchy crust, sugary tomato sauce, and a little cheese is a so-called high fat meal.

    I agree with Olivia: let’s see results of a carb-only test and a fat-only test. If you want to know your own BG test results, you can do it with a $10 meter at Walgreens.

  • @ Guru & Jonathan,

    I also have a feeling that insulin alone is not to blame for everything. The insulin index shows that beef gives a higher insulin response that oats and that dairy yogurt gives a higher insulin response than bread.

    I have said before that low carb is just as simplistic a view as low fat as a solution to obesity. Maybe the emphasis should a high vegetable diet (low carb veggies of course) with fat and protein and carbs playing a minor role? Im always on the lookout for new theories.

  • Jonathan

    Insulin varies all the time. http://en.wikipedia.org/wiki/Insulin_oscillations
    It’s one of those hormones that we can not live without. We can’t prevent it’s release and we don’t want to because we need it to store fuel and nutrients for later use. Now, keeping it from being chronically elevated is another story.

    I can’t see the original curves the numbers came from but just looking at the numbers, meat really wasn’t high on the index in my opinion.

    Someone wants to study the index more that would be fine but they need to do it well. I would think it more important to measure the insulin levels, blood sugar levels, and blood free fatty-acid levels over a 6 hour period; all along with general hunger and energy levels. They may also want to feed/test specific macronutrients instead of foods.

    A high veg diet? That would just be a gut filler. If that satisfies you then fine, but we don’t have the guts of a gorilla to handle all that so we would just be a “Salad Shooter” later. What does digest would be the soluble fiber being converted into fatty acids by gut flora.

  • Mark91345

    I am familiar with the Insulin Index and like the concept of having foods indexed this way, as there were some “surprises”, like white bread and whole-wheat bread have about the same insulin index; a potato or baked beans are 20% worse than white bread; peanuts have very low insulin response.

    The problem is there are so few items that were tested, but it gives an idea about how insulin responds. I wish there were more items tested, but such is not the case

  • Pizza, a high fat food? Sure. Other foods that are touted as “high-fat”: french fries, milk shakes, donuts, peanut butter, non-skim milk, etc. A cavalcade of sugar and starch. Pahhhleeze.

    Hmmm… french fries, milk shakes, donuts … I’m hungry.


  • Karen LaVine

    Hi Jimmy:

    To keep things completely clear: fats do NOT directly raise blood glucose levels. But certain types of dietary fatty acids DO cause hepatic insulin resistance, and that, when one has diabetes, can measurably raise basal glucose levels.

    Re: Jonathan’s statement that insulin levels can’t get too high or BG will get too low – he’s forgetting glucagon – the body can produce more glucagon to counteract the effect of insulin, so people can have high insulin levels without a low BG – that just means they are also hyperglucagonemic. And don’t forget the importance of leptin – there is an interesting clinical trial going on now with people with type 1 taking basal insulin and a fixed dose of leptin twice a day – this combo may be more effective than insulin alone in controlling glucose levels, and appears to require significantly lower doses of insulin which are closely correlated with improved health outcomes.

    Here’s a link to an article that further delineates what I have been observing about dietary fats. This was written by Gary Scheiner, MS, CDE. Gary recently made an excellent presentation for the American Association of Diabetes Educators re: evaluating continuous glucose monitoring data. During this presentation, Gary briefly alluded to the importance of evaluating the effect of other macronutrients besides carbs on subsequent BG patterns in people with diabetes. Through David Mendosa, I asked Gary to write an article describing exactly what he has been finding. His article completely reinforces what I have been observing, and I’m confident that many other CDE’s have also been seeing a similar effect.
    The upshot of the article: After reviewing many continuous glucose readings in many different people with diabetes, Gary has found that post prandial BG rises 4-10 hours after a meal high in fat, long after even slow carbs have been digested. He has developed specific insulin regimens to deal with this.

    And this review of recent research provides one specific explanation re: why this is occurring.

    Therefore, dietary saturated fats, by triggering hepatic inflammation, insulin resistance, elevating insulin levels and increasing signaling in the I/IGF pathways, tip the metabolic balance away from survival/repair pathways to reproductive pathways. This appears to correlate with a decreased probability of an extended lifespan.

    It is interesting to see, when monitoring people with diabetes, that their fasting blood glucose can come down significantly from eating just one meal lower in fat with low to minimally processed carbs the evening before. This indicates their hepatic insulin resistance/sensitivity shifts rapidly.

    Jonathan does mention the saturated fatty acids in fiber that improve peripheral insulin sensitivity. Butyric acid is a very short chain saturated fatty acid. When we eat dietary fiber, gut bacteria convert some of that dietary fiber into butyric acid. We absorb it and it improves our peripheral insulin sensitivity. He’s right: We can get very healthy dietary insulin-sensitizing fat by eating high fiber foods. And there is also stearic acid, another saturated fatty acid which also seems to have some health benefits.

    But it is difficult to divide out the inflammatory saturated fats and omega 6’s from other dietary fats, so to lower my insulin levels while simultaneously controlling my blood glucose levels, I eat a very low glycemic index diet (NO processed fructose/sucrose, etc so triglyceride levels stay very low, and lots of peripheral insulin-sensitizing dietary fiber from MINMALLY PROCESSED WHOLE grains) and a very low fat intake (to limit both saturated fats and omega 6’s) thereby decreasing inflammation, optimizing hepatic insulin sensitivity and decreasing basal insulin levels. (And I take algae-derived EPA/DHA to help keep immune system inflammatory responses low.)

    Yes, Jimmy, this diet does “float my boat” and I wonder who else it works for as well. (During his interview, Gary T mentioned some participants of specific low carb studies who just couldn’t lose weight despite following very low carb diets. I wonder if they had elevated inflammatory markers and increased basal insulin levels?) I used to be on a low carb diet, and I have never been overweight, but on this low fat, low glycemic index diet, my weight is actually a couple of pounds lower now than it’s ever been in my adult life: BMI 19.6 and my fasting glucose levels range a little lower than when I was low carbing. I believe that by minimizing dietary factors that cause inflammation and insulin resistance and maximizing factors that decrease inflammation and increase insulin sensitivity, I am increasing my chances for a long health span. And I don’t get hungry – I eat large amounts of fresh veggies, (lots of them straight out of my garden) cooked barley, quinoa, buckwheat, oat groats, etc and some legumes every day. And I also exercise – I follow a regimen that research has shown improves… insulin sensitivity!

    From your October 27th post about Cynthia Kenyon’s research:

    They also accurately report that consuming more carbs which raises insulin also leads to an increase in cholesterol, raises blood pressure levels, and releases triglycerides into the blood stream to significantly boost the chances of getting heart disease.

    Well, Jimmy, that’s sure not true for me! On this very low glycemic index and low fat diet, my fasting c-peptide level last year was BELOW normal range, (0.6 ng/ml – range 1.1-4.4 ng/ml and my insulin level just came out as <2.0 uIU/mL range 0.0-29.1 because that lab doesn't measure below 2.0) and yet my fasting BG @ 83 mg/dL and A1c 5.7% were normal for a 55 yr old woman. I'm thinking that this diet helps optimize not only my insulin sensitivity, but also my leptin sensitivity, and this leptin sensitivity helped my glucose stay in range despite subnormal levels of basal insulin.

    In order to evaluate your insulin sensitivity/resistance if you are not taking insulin, a serum fasting insulin (or c-peptide) and fasting glucose level have to be drawn. This allows you to calculate your insulin sensitivity, via HOMA2. If you do get a chance to interview Dr Kenyon, if possible, please ask her what her fasting insulin (or c-peptide) and glucose levels are.

    I'm looking for other people who monitor their fasting insulin and glucose levels to learn what they find helps them optimize their insulin sensitivity.

    Thanks so much,

    P.S. Sorry this is so long, you can cut and paste this as you please – I know you're a fellow traveler searching for the truth, wherever it may lie.

  • Michael Scott

    I have learned from experience never to question Gary Taubes! Very few people have Gary’s knowledge and the subject of weight loss. Keep up the great work Gary!

  • Lillian

    I read Gary Taubes book, Good Calories, Bad Calories, and it really has help to change my life. Losing weight in an easy and steady manner. I have been on low carb diets in the past. Both times they worked, but both times I came off them because of health lies. I did not come off them because low carb is hard to stick to. I am not hungry. I am not shaky. I am not waking up several times in the night, from hunger. I am not trying to get back to sleep hungry.
       I am eating in the morning, but I don’t wake up starving. I enjoy all the food I am eating. The only problem. I can be enjoying a really fantastic meal, when my stomach says, No More. On a high carb diet, my brain pushed past the stomach, and I ate. On a low carb diet, my stomach wins. The only reason I ever left low carb diets in the past was because of presure from well meaning friends and my doctor who lectured me after I lost 20 pounds on a low carb diet.
        When they say that people on low carb gain weight back as much as low fat diets, have they taken the social pressure into account. No one ever interveins to push you off a low fat diet. Thanks to Gary’s book, it will never happen to me again.

  • dr robert peers

    very interesting argument…  trouble is, some excellent detective work in 1935, by professor sir harry himsworth, showed a very strong link between rising diabetes mortality and rising dietary fat intake in western nations, from 1904 (with a transient fall in such mortality in fat-starved germany, in the first world war)…furthermore, diabetics ate more fat, and LESS carb during this emerging diabetes epidemic…high protein diets may help weight loss because protein satisfies quickly, so you eat less…fatty diet will not elevate blood glucose immediately…but it will soon cause faulty cell membrane structure and insulin resistance…glucose levels are then controlled by gradually rising insulin levels, which eventually decline, as oxidative damage hits the beta cells, so allowing glucose to get out of control, after many years of fatty diet-induced insulin resistance and oxidative stress..the disease is called type 2 diabetes, but should be called dietary fat toxicity…simple carbs may indeed increase triglycerides, but raised trigs alone do not affect heart risk–and complex carbs do not have that effect at all, so why avoid them too?…postprandial rises in glucose and trigs after simple-carb foods are only seen in certain people, who may be anxious [due to FATTY maternal diet–e.sullivan, 2010], and anxiety inhibits insulin secretion [ha ha]…if i was gary taubes’s doctor, i would ask him how he cooks his red meat (grilling, frying and barbecuing cause cancer-promoting chemical changes in the meat surface (see SINHA R AND RED MEAT, on PubMed)…I would also ask him why eaters of complex carbs have so much energy, and why most healthy centenarians are grain and legume eaters—all seed foods are rich in inositol, which mimics the genetic effects of caloric restriction (J Barger, 2008), and which also activates the master energy gene pgc-1-alpha….also, pizza ham is nitrite-treated, and nitrites are pro-carcinogens, and may be toxic to beta cells as well