Saturday, January 30, 2016

Testing the Insulin Model: A Response to Dr. Ludwig

Dr. David Ludwig, MD, recently published a response to my critique of the carbohydrate-insulin-obesity hypothesis.  This is good because he defends the idea in more detail than I've encountered in other written works.  In fact, his piece is the most scientifically persuasive defense of the idea I can recall.

Before we dig in, I want to emphasize that this is science, not tribal warfare.  The goal is to arrive at the best answer, rather than to win an argument.  I'm proceeding in good faith, based on my belief that Ludwig and I are both serious people who care about science and human health, and I hope my audience will do the same.  That said, let's get to it.

Introduction: CICO vs. insulin.  Or is there a third model?

Similar to Gary Taubes, Ludwig presents a choice between the calories-in, calories-out model (CICO) of obesity, and the insulin model.  The CICO model is the idea that our body weight is determined by voluntary decisions about how much we eat and move, and in order to control our body weight, all we need is a little advice about how many calories to eat and burn, and a little willpower.  The primary defining feature of this model is that it assumes that food intake and body fatness are not regulated.  This model seems to exist mostly to make lean people feel smug, since it attributes their leanness entirely to wise voluntary decisions and a strong character.

I think at this point, few people in the research world believe the CICO model.  Ludwig and I both agree that it provides a poor fit for the evidence.  As an alternative, Ludwig proposes the insulin model, which states that the primary cause of obesity is excessive insulin action on fat cells, which in turn is caused principally by rapidly-digesting carbohydrate.  According to this model, too much insulin reduces blood levels of glucose and fatty acids (the two primary circulating metabolic fuels), simultaneously leading to hunger, fatigue, and fat gain.  Overeating is caused by a kind of "internal starvation".  There are other versions of the insulin model, but this is the one advocated by Ludwig (and Taubes), so it will be my focus.

But there's a third model, not mentioned by Ludwig or Taubes, which is the one that predominates in my field.  It acknowledges the fact that body weight is regulated, but the regulation happens in the brain, in response to signals from the body that indicate its energy status.  Chief among these signals is the hormone leptin, but many others play a role (insulin, ghrelin, glucagon, CCK, GLP-1, glucose, amino acids, etc.).

Here is a simplified schematic overview of how the system works, from a 2012 review paper I wrote with my scientific mentor Mike Schwartz, titled "Regulation of food intake, energy balance, and body fat mass" (1).  This figure summarizes more than a century of research in our field:

Here's the gist of it: there are negative feedback loops between the brain and fat tissue, and between the brain and the gut.  These are what regulate body fatness and appetite.  The primary known feedback signal that regulates body fatness is leptin-- a fact that has remained scientifically unchallenged since shortly after its identification in 1994.  Insulin plays a role as well, acting directly on the brain in a manner similar to leptin, although much less powerfully.  As you can see, this model doesn't resemble the CICO model-- or the insulin model.

Regulation happens principally as a result of the brain changing the number of calories entering and leaving the body (in humans, mostly entering)-- so the much-maligned calorie maintains a central role in the process.  Even though calories aren't the first link in the causal chain, they are nevertheless a critical link.

Most people in my field also believe that calorie intake is determined both by hunger (homeostatic eating), and factors other than hunger (non-homeostatic eating).  I agree with them.

So this third model acknowledges the importance of regulation, the importance of unregulated factors, and the importance of calories.  Although it's more complex than the insulin and CICO models, it has the advantage of being supported by a large body of evidence.  Let's call it the leptin model, for simplicity's sake.  I won't spend much more time on it because it's peripheral to this debate, but I did want to articulate it as a third model so we aren't limiting our choices to 1) a CICO model that is obviously wrong, and 2) the insulin model.

I'm not going to do a point-by-point response to Ludwig's point-by-point response, because that would be exceedingly cumbersome.  What I'll do instead is try to distill out the most important concepts, and discuss those.  In particular, I'll be focusing on key lines of evidence that have the potential to support-- or falsify-- the insulin model.  This is where the rubber meets the road for hypothesis testing.

Is obesity caused by internal starvation?

In my initial post, I pointed out that people with obesity have normal or elevated levels of blood glucose and fatty acids-- not lower levels as the insulin model predicts (1a, 1b, 1c).  This suggests that 1) they are not suffering from "internal starvation", 2) they are not suffering from excessive insulin action (including on fat cells), and 3) therefore this mechanism cannot explain the well-established fact that people with obesity tend to eat about 20 percent more calories than lean people (2, 3).

Ludwig countered that "static analysis after obesity has developed, rather than during its dynamic stages, are misleading".  In other words, I was focused on the wrong phase of obesity; we should be looking at the weight gain phase instead.  Let's think through the implications of this statement.  It implies that internal starvation makes us overeat as we're gaining weight, but doesn't explain continued overeating once we reach the obese state, or the fact that the obese state persists after internal starvation is no longer occurring.  It therefore also requires the existence of two separate mechanisms that cause overeating, one involving internal starvation and one not, and requires a hand-off between them at some point in time.  I'm not aware of experimental evidence that supports any of this.

Let's take a closer look at the weight gain phase, as suggested.  Is there evidence that people who are gaining weight have lower glucose and fatty acid levels than people who aren't?  There is not.  I'm not aware of any evidence in humans or in animal models of diet-induced obesity that circulating glucose or fatty acid levels are lower during active weight gain.  In Mike Schwartz's lab, we have done countless glucose measurements in mice and rats gaining weight rapidly due to a fattening diet, and we have also measured circulating fatty acid levels at times, and neither one is decreased during active fat gain (3a).

Ludwig acknowledges this lack of evidence in his article, but instead focuses on indirect evidence that's not particularly relevant to common obesity, such as hypothalamic lesions in rodents (As an aside, Bruce King has shown that the fattening effect of VMN lesions does not depend on increased insulin levels (4), and John Brobeck showed in the 1940s that these lesions cause voracious overeating as soon as the anesthesia wears off, not after some period of altered fat tissue metabolism (5).  In any case, alterations in fat tissue metabolism are the result of the brain lesion, showing that the brain is influencing fat tissue metabolism in these experiments, not the reverse).

So in summary, there is no direct evidence that internal starvation occurs during 1) weight gain or 2) the overweight/obese state, and there is some direct evidence that it does not occur.  The internal starvation hypothesis rests entirely on indirect evidence that I don't find particularly pertinent.

But let's take a step back for a minute.  Are low circulating fatty acid levels able to activate the brain's hunger circuits in the first place, leading to increased food intake and weight gain?  Because if they aren't, then it's hard to understand how insulin (or anything else) would lead to overeating by reducing fatty acid levels.  This is testable.  To test it, we would want to reduce circulating fatty acid levels and see if it provokes a hunger response, and ideally, weight gain.  Fortunately for us, this study has been done.

Hideo Makimura and colleagues recently published the results of a 6-month randomized, placebo-controlled study in which they used a drug called acipimox to chronically reduce circulating free fatty acid levels in people with obesity (6).  The drug caused a substantial (38%) reduction of free fatty acid levels for the duration of the 6-month study.

According the idea of internal starvation, the acipimox group should have been ravenously hungry and gained fat rapidly, and energy expenditure should have declined as well.  Yet the researchers reported that "caloric and relative macronutrient intake did not change significantly between groups".  Neither group's BMI or body composition changed significantly, nor did their energy expenditure.  The study was published in the Journal of Clinical Endocrinology and Metabolism, a respected journal in my field.

The brain monitors a number of signals from the body to measure energy status, and it uses these to set your appetite.  Apparently, low circulating fatty acid levels are not one of the signals the human brain pays attention to when it sets appetite.  This seriously undermines the concept of internal starvation, and consequently the insulin model.

In contrast, we have strong evidence that the human brain pays close attention to circulating levels of leptin.  When leptin levels are low, whether due to weight loss or mutations in the leptin gene, it triggers a full-blown starvation response including increased hunger, increased responsiveness to food cues, and reduced energy expenditure.  Bringing leptin levels back to normal via injection terminates this starvation response, clearly demonstrating that low leptin was the principal cause (7, 8, 9, 10, 11, 12, 13).  These results don't leave much room for insulin or fat cell metabolism to be involved in the body's starvation response (except perhaps as downstream effectors of leptin).

Do high insulin levels predict fat gain?

This is a straightforward prediction of the insulin model, and we have many human studies with which to test it.  In my initial post, I cited a systematic review paper that summarized the results of 22 prospective studies examining this prediction (14).  Overall, the literature suggests that people with high insulin levels do not gain more weight or fat over time than people with low insulin levels.

Ludwig countered by arguing that what matters isn't insulin levels, but insulin action; in other words, how much of an impact the insulin actually has on cells and tissues.  Insulin action is determined both by insulin levels and insulin sensitivity (how well tissues "hear" the insulin signal).  This is the same argument I have used to question the insulin model, because proponents often assume that high insulin levels automatically imply elevated insulin action.  They generally overlook the fact that people with high insulin are almost invariably insulin resistant (including their fat tissue), and that insulin action is normal or reduced (as judged by normal or elevated blood glucose and fatty acids, and normal or elevated whole-body lipolysis rates).

In fact, this is precisely the reason why insulin injections, insulin-producing tumors, and similar non-physiological examples cited by Ludwig and Taubes are irrelevant to the question at hand.  These arguments attempt to draw an analogy between non-physiological insulin excess and garden-variety hyperinsulinemia-- but in the latter case, there is no evidence that insulin action on fat cells is actually increased, and a fair bit of evidence that it isn't.

In any case, I agree with Ludwig's basic premise that insulin action is what really matters.  Yet if we return to the review paper I cited, it didn't just consider studies that measured fasting insulin levels.  The studies measured multiple insulin-related variables, including fasting insulin levels, insulin sensitivity, and the insulin response to a glucose challenge.  Although we could cite individual studies to support any hypothesis we want, the overall literature suggests that none of these variables are reliably associated with weight gain.  In fact, I'm not aware of any insulin-related variable that is reliably associated with weight or fat gain in humans, despite intensive research in this area.  Perhaps we're not measuring the right thing, or not measuring in the right way, but that possibility is nothing more than speculation at this point.

Ludwig suggests that we should be looking for the effects of insulin during the weight gain phase.  That is exactly what these studies have done, and they did not find that weight gain was consistently associated with a distinct insulin signaling profile in any measured way.

So we've measured insulin levels, insulin sensitivity, and the insulin response to carbohydrate.  We've measured them in people who are gaining weight and people who aren't.  We've measured them in multiple races and ages.  And no clear pattern has emerged suggesting that insulin signaling might be playing an important role in the fattening process.  Again, it's always possible that we haven't measured the right variable yet, or haven't measured it in the right way, but that is nothing more than speculation at this point.  Currently I find this evidence rather difficult to reconcile with the insulin model.

Do low-glycemic diets work for fat loss?

If rapidly-digesting carbohydrate that markedly elevates insulin levels is a major driver of overeating and weight/fat gain, then low-glycemic diets that reduce insulin exposure should be an effective tool for reducing food intake and body weight.  In my initial post, I reviewed evidence suggesting that 1) the glycemic index doesn't reliably predict the satiety response to common foods, and 2) low-glycemic diets are ineffective for weight control.

To the first point, Ludwig countered that the study I cited only measured the satiety response over a two-hour period, but hunger usually occurs later (15).  He also cited a review paper he wrote in 2002 suggesting that single-meal studies have generally found that low-glycemic meals are more sating than high-glycemic meals (16).  This is a fair point, and the cited evidence does support it (although many of these studies were not controlled for other differences known to affect satiety, such as fiber content, palatability, and calorie density).  However, findings in this area haven't been especially consistent, and in a few cases high-glycemic foods were actually more sating (17, 18, 19).  I do acknowledge that the weight of the evidence is in his camp on this point.

But in any case, what we really care about is not what happens at a single meal, but what happens over the long haul.  That is the question my second point addresses.  We have quite a bit of evidence suggesting, fairly consistently, that low-glycemic diets don't work for weight loss (20, 21, 22, 23, 24, 25, 26).

Ludwig countered that these studies "suffer from severe non-compliance, limiting inferences".  In other words, people don't follow the experimental diet very faithfully, explaining why the diets appear ineffective.  While I agree that adherence is always a limiting factor in human trials that don't lock people up in a research ward, poor adherence doesn't explain the null results in this case.

For one thing, adherence is always a problem in human diet trials, but they nevertheless often report weight loss.  This has been shown for the low-carb diet, the Paleo diet, the vegan diet, the Mediterranean diet, simple portion control, and even the embattled low-fat diet.  Some of these are tough diets that people have a hard time sticking with, but they still cause measurable weight loss despite imperfect adherence.  In diet trials, adherence rates to low-glycemic diets are similar to those of other diets, so it's hard to understand how this would explain the lack of efficacy.

But let's get specific.  I'd like to start with a study that may be the purest test of the hypothesis available (27).  It was conducted by Walter Willett and Frank Hu, two of Ludwig's colleagues at Harvard.

They randomized 203 healthy women (average BMI = 27) to one of two diets: high-glycemic or low-glycemic.  Differences in glycemic index were achieved primarily by eating different types of rice that have different digestion speeds, so the study was relatively well controlled for other aspects of diet like fiber, calorie density, palatability, etc.  The overall glycemic index (and glycemic load) of the diets differed by two-fold (40 vs. 79), which implies a large difference in both glucose and insulin exposure.  This large difference was sustained for the full 18 months of the intervention.

There were no significant differences in hunger or calorie intake between the two groups.  At two months, there was a trivial difference in body weight of 0.4 kg favoring the low-glycemic group, but this disappeared by the end of the study.  The conclusion is that large and sustained differences in post-meal glucose and insulin exposure have no meaningful long-term impact on hunger, food intake, or body weight in women (who were overweight on average).

Let's consider another study-- the one from Ludwig's group that I cited previously (28).  This 18-month randomized trial pitted a low-glycemic-load diet against a low-fat diet in 73 obese young adults.  The low-glycemic diet was focused on whole foods like vegetables, beans, and fruit, at the expense of refined grains, starchy vegetables, fruit juice, and sweets.  The low-fat diet was focused on low-fat grains, vegetables, fruits, and beans, at the expense of of added fats, sweets, and high-fat snacks.  The low-glycemic diet was a bit lower in carbohydrate (40% vs 55%), and both diets were fairly high in protein (25%).  Neither diet was calorie restricted.

Glycemic load differed substantially on the two diets, and a difference was maintained over the 18-month study, although it narrowed at 12 and 18 months.  This implies substantial differences in glucose and insulin exposure.

So what happened?  Hunger, calorie intake, and participant satisfaction didn't differ between groups.  Both groups lost weight (4-5 kg at max), but the magnitude and trajectory of weight loss between the two groups was virtually identical (see graph below).  Substantial differences in diet-related glucose and insulin exposure did not result in detectable differences in hunger, calorie intake, weight loss, or weight regain.
18-month weight changes in Ebbeling et al. Note the p-value of 0.99,
indicating that the two trajectories are statistically indistinguishable.
In his response to my post, Ludwig pointed out that subjects who had a greater insulin response to a glucose challenge at baseline experienced more weight loss on the low-glycemic-load diet.  This is true, but we have to keep the finding in context.  The study wasn't randomized to answer this question (unlike a recent study from Christopher Gardner's group; 29), so this is what's called an "exploratory analysis".  These carry less weight than the primary outcomes of a study, which are those that I discussed above.  Exploratory analyses can certainly be informative, and there's nothing wrong with conducting and publishing them, but they are better suited for generating hypotheses than for testing hypotheses.  And they never trump the primary outcomes of a study.

The most straightforward interpretation of the study is that if you're a young adult with obesity, reducing the glycemic load of your diet, and consequently your insulin exposure, has no special ability to curtail your appetite or help you lose weight.  You might as well go on a low-fat, high-carbohydrate, high-glycemic diet.  This is consistent with the results of the previous study I discussed, and the rest of the literature as a whole.

It remains possible that low-glycemic diets are helpful for a subset of people, but current evidence suggests they are not very effective as a general strategy for managing appetite and weight-- counter to what the insulin model predicts.


In summary, several key predictions of the insulin model are not supported by the evidence, explaining why this model doesn't get much traction in my field.  There is essentially no direct evidence that the proposed mechanism occurs during or after normal weight gain, a fair amount of direct evidence that it doesn't, and the arguments in favor of it are based on indirect evidence whose relevance to common obesity is often questionable.

I thank Dr. Ludwig for the opportunity to clarify my thoughts on this, and I wish him luck in his ongoing research.


Unknown said...

Hi Stephan,
Re:"Apparently, low circulating fatty acid levels are not one of the signals the human brain pays attention to when it sets appetite."
Have you encountered if this is also true during a diet induced ketosis or only if the metabolism is running mainly on glucose?

Bill Shay said...

I am not a scientist but do have an interest in this subject. I find your article, along with others you have written and posted here, to be convincing. So given the lack of scientific support for the CICO and insulin models, we are left with what you call the leptin model. This leaves the question of how does the leptin hormone work, what factors influence it, and what can we humans do in light of this to lose fat and improve health? I hope you follow up this article and shed more light on these questions, or point us to other relevant articles. Your article leaves me "hungry" for more. Thank you.

karl said...

"low-glycemic diets are ineffective for weight control."

A moot point? - Why would they work? They would tend to keep a minimal level of insulin up long term - preventing the combination of low insulin sensitivity and low insulin needed for fats to leave adipose tissue.

"Chief among these signals is the hormone leptin, but many others play a role (insulin, ghrelin, glucagon, CCK, GLP-1, glucose, amino acids, etc.)"

Why is leptin "Chief"? What about ghrelin, Adiponectin, seritonin, thyroids, ephrinephrin, melatonin, glucocorticoids, estrogen, xenoestrogens? Leptin is short term control has a circadic cycle - the problem is long term. If leptin is key and blocked by elevated trygly (as you early wrote) - it would explain why both fructose sugars and high carb (both increase trygly) can effect weight - yet the explanation fails to address what gets broken.

I don't think there is one stimulus for hunger - or even one sensation. Subjectively for me there is a great difference in hunger following a carb-crash than what follows a long day of physical activity or what comes before breakfast.

But here is the real rub - there are many people that have normal weight until their 30's or more. Their self discipline has not changed. Yet, suddenly they are gaining weight, failing at their attempts to diet. Something has clearly changed - something broke. When attempting to diet, people are miserable - lack energy, loose muscle mass. I don't see how lack of some kind of puritanical mindset can explain these complications.

This is not a simple feedback loop system - instead you have a mass of feedback looks - some nested some in series some in parallel. And sadly the loops are not just proportional - they are at once non-linear - some integrate some differentiate. The post-synaptic receptors in the brain can down-regulate (papers show the change in gene expression). This is a long term problem that could be quite confounded with short term signals. thus:

"people with obesity have normal or elevated levels of blood glucose and fatty acids-- not lower levels as the insulin model predicts" Fails to make a point - it shows that the body is keeping things in normal ranges - what counts is the amount of fat stored and energy taken out of easy access. If too much fat is stored - or retained - the body will want/need to get that energy elsewhere - I could imagine higher BG - not lower as consumption is increased.

LC diets might also work via their effect on Adiponectin - or reduction of interleukin.

Either idea: that it is simply insulin or insulin has nothing to do with it are obviously both wrong. It misses the understanding that biology is amazingly complex SYSTEM. Insulin and glucose control obviously get broken - buy is that causative or a secondary symptom?

Attempts to understand this by over simplification will probably fail - the key feedback signal may not vary much at all - yet they can be key bit that is failing.

Puddleg said...

A couple of quick points - the GI index is a pretty poor predictor of the individual response to a food.

On the "internal starvation" question, has anyone considered the importance of glucose and lipid cycling in metabolic diseases?
If portal glucose is high and the liver is insulin resistant, some of the oxaloacetate from glucose will be used to make more glucose, similarly, acetyl-CoA from fatty acids will be cycled into other lipids. This means that the liver requires more substrate to generate the same amount of net ATP, and puts more substrate into circulation. There is an inflationary process at work; more substrate is available, but it is not worth as much.
Starvation could be a relative term, and these substrates are easily stored as fat in a high-insulin milieu if the adipocytes are less insulin-resistant than the liver.

Audacity17 said...

Some time ago I found that drinking sugary Coke(instead of diet) at the buffet resulted in me eating less food.

thhq said...

I'm also interested in how to regulate leptin's hunger signaling. Any information on how to stop the clarion call of hunger at 4 AM would be helpful. It hasn't changed since I finished losing a significant amount of weight 9 years ago.

I found that the CICO model worked for weight loss when I tailored to my metabolic rate. I keep a reasonably accurate log of eating and exercise, and back calculated a personal BMR from my rate of weight loss 9 years ago. Having done it, I feel that you have to be losing weight at a fairly substantial rate - 5-10 lbs per month - and you have to have at least a couple of months of steady loss, to get an accurate estimate. My personal experience both losing and maintaining was (and still is) a metabolic rate very close to what Harris Benedict predicts. I adjust this number with age and weight changes using HB, count both food and activity, and maintain on a 100-200 calories per day deficit.

Anonymous said...

Just curious, Dr Guyenet. You're clear that you don't think insulin is relevant, but at the same time you're aware of the success of low-carb diets. You seem to lump them in with Paleo, and I assume you think the secret sauce is the lack of hyperpalatability.

Studies have shown that ketosis is associated with suppression of ghrelin, and that levels go back up when out of ketosis.

Ad lib low-carb feeding studies show that participants voluntarily eat at a significant calorie deficit, and anecdotal evidence of such is widespread on low-carb forums.

So would you agree that low-carb appears to have the highest success rate for weight loss, and if so, how would you explain that success? If due to suppressed ghrelin, why would the body dampen hunger signals when carbs are restricted?

Stephan Guyenet said...

Hi Normalcarb,

I wouldn't go so far as to confidently say insulin isn't relevant. It's always a matter of probability in science. There are many versions of the insulin hypothesis, and this post addressed one of them that I think is likely to be incorrect. I don't find any of them compelling at this point, but in some cases that's simply because there isn't much evidence available to evaluate them. I can't predict where the evidence will go, and although I think it's unlikely, I'm prepared to change my mind about insulin if research clearly supports some version of the hypothesis.

For example, James Johnson at UBC is doing mechanistic research on the role of insulin in obesity in rodents. He and I disagree on how to interpret his published data, but he is doing legitimately interesting work that has the potential to shed new light on this question (as an example of how he's thinking outside the "carb-insulin box", he doesn't assume that carbohydrate is what causes hyperinsulinemia-- he's using a high-fat diet model of hyperinsulinemia in his work). Who knows what will come of this research.

All this is to say that I'm trying to maintain a humble attitude about it, because I think acknowledging the limited state of your own knowledge and the uncertainty of the world leads to being right more often in the long run.

Any hypothesis that wants to explain obesity would have to tie in with the leptin-hypothalamus feedback loop somehow, because we know that's the system that defends the lower bound of adiposity and therefore maintains the obese state. Basically, it would have to explain why the leptin 'setpoint' is higher. For example, Robert Lustig's hypothesis is that hyperinsulinemia causes leptin resistance in the hypothalamus. There is no supporting evidence for this mechanism, but it at least ties in to the system that we know regulates body fatness, making it much more plausible on its face.

Unknown said...

Interesting article, thanks. Perhaps the GI / GL of the food isn't mirrored by the individual's response to it. So, the high GI food for John isn't a high GI food for Tim; Personalized Nutrition by Prediction of Glycemic Responses

JasmineJohend said...

Dr Guyenet do you have any thoughts on increasing leptin levels after weight loss to stop the rebound hunger? Andro from Suppversity wrote an article on Baiba, and a Google search found interesting info and some studies about how it affects leptin and beige fat. It doesn't seem to have any evidence for actual weight loss but more for prevention of weight gain. I would love to know your thoughts.

Joe said...

Continuing on normalcarb's quiestion, and long-term compliance aside, since all diets have poor long-term compliance.. There's a couple dozen studies all showing greater weight loss on low-carb diets. Ad-lib LC vs calorie-counted low-fat, the deck is massively stacked against LC, yet it still wins.. I find this incredibly interesting.

What is it about low-carb (real LC diets) that cause spontaneous reduction of food intake? LC tends to be calorically dense too, which should make them easier to overeat on, yet that's not what tends to happen.

Isn't this an important clue as to what's happening? How does this fit into the leptin model?

I think it would be interesting to see if low-carb can be broken, eg introduce something into it that causes overeating and obesity while still maintaining carb intake to very low levels. It'd be another important clue.

thhq said...

I did a little further reading on the interactions between leptin, insulin and melatonin; particularly in regards to weight loss maintenance. Weight loss and ageing depress leptin levels, which would explain my ongoing hunger/low sating in defending a 20% reduction in BMI. Increasing melatonin reportedly depresses insulin and increases leptin (in conjunction with insulin). So I've constructed a simple N=1 experiment using readily available melatonin: eat a 500 mg melatonin tablet ahead of each meal. I expect this to result in better postpradial sating due to higher bloodstream leptin (resulting in less after-meal snacking to satisfy hunger), and possibly a shift to favor fat over carb metabolism due to lower bloodstream insulin (potentially resulting in weight loss).

But as with many N=1 tests in the past, there's probably some flaw in the plan...

thhq said...

Oops 500 mcg not 500 mg.

karl said...

@Stephan - "All this is to say that I'm trying to maintain a humble attitude about it, because I think acknowledging the limited state of your own knowledge and the uncertainty of the world leads to being right more often in the long run."

Glad to hear you say that. There has been a really very bad track record in medicine of treating and advising people based of speculative theories. (When your doctor says he doesn't know - you might have a really good one. )

I don't like to use analogies, (they often mislead), but this might help others understand why there isn't a simple answer. Many years ago - on color TV sets I used to do alignments - there were several controls and ALL the controls influenced the other controls - to make the adjustment correct, one had to adjust the controls in the correct order and repeat the process a few times. The point is the system worked - but everything interacts. All the settings changed, but knowing which parts drifted in value was not apparent.

That BG, insulin, leptin, ghrelin, glucagon etc etc - all interact - changing one effects the others (sometimes very indirectly ) - non linearly, sometime integrating sometimes, pulsatile, (Normally, insulin is secreted from the pancreas in pulses into the portal vein) means that figuring out what is causative rather than correlative is very hard to do. Poorly designed studies muddy-the-water and consume limited research resources. ( The lack of synthetic diet studies is mind-boggling) - biologists poor understanding complex feedback loops further muddies-the-water (not totally their fault - the math is complex and dry - computer models can often mislead more than help). Dis-regulation of appetite of only 1% can cause obesity - the signal may be very small.

There is a serious pandemic of T2D - people are suffering, loosing quality of life. I think there is likely a factor in diet that is responsible. My list of probably cause has changed over time - and continues to be speculative.

- PUFA - particularly linolenic acid - from concentrated veg oils (not a natural human food)
- Fructose containing sugars - ( only in our diets after 1600s in any large amount)
- Excess carbohydrates - ( there may be a variation of evolutionary adaption - wide spread only after farming.
- xenoestrogens

Some want to say it is all about eating too much - (bad ethics?) but I know too many highly motivated people that over-exersize and fail, diet and fail, pop pills and fail. Something has broken in their bodies and their doctors can't really help them much at all.

My list of possible broken bits:

mDNA damage
Hypothalamus damage
Pancreas damage
Adipose damage.
liver damage

There is the case of progressive lipodystrophy - picture

Thin on top - seriously fat on the bottom. Same hormones in the blood stream - but could it be a break in a nerve? Autonomic nervous system? Can we make an experiment to stimulate nerves that feed adipose tissue? Could it be a form of neuropathy - similar to peripheral neuropathy which is often co-morbid in T2D?

Unknown said...

These posts are excellent. Really appreciate your perspective on this topic.

I always wonder why the insulin hypothesis has to be fat cell centric. I think it would make more sense if it conceded the role of the brain - especially since the brain itself can become insulin resistant which could lead to missing satiety cues and over eating:

Aegirsson said...


You only look at nutrition. The other part of the equation (and in my opinion a rather large elephant in the room) is our ridiculously sedentary lifestyle. I don't mean for myself, I am very active and intend to remain so. But most people don't move nearly enough. Even going to the gym does not always cut it: I see people going to the gym by car, lifting ridiculously small weights at high reps, using guiding machines instead of free weight lifting, or on the other hand, climbing on a bike or shuffling on the treadmill for 20mn. And that's it. Shower, back to the car, home. I once did a rough statistics based on my own observations at work and around where I live. About 8 out 10 men above 35-40 yo have a pot belly. maybe 5 out 10 are already showing signs of developing one during their late 20's. This is serious shit.

Anonymous said...

@Joe, I took Guyenet's answer to suggest that leptin is king, so any model has to include leptin.

I believe Ludwig has directly studied the effects of diet on leptin. He looked at high-GI vs low-GI vs low-carb, and found that low-carb reduced leptin the least of the three. He also suggests that the overweight have leptin resistance and that a low-carb diet somehow fixes that.

So low-carb may work by improving leptin sensitivity and mitigating the reduction in leptin secretion. That could explain the reduction in appetite.

thhq said...

Weight loss is known to reduce both insulin and leptin resistance.

Ergo increased activity and reduced caloric intake are important. They're the elephant in the room.

Sugar and salt are a sideshow. Magic tricks sell tickets though.

karl said...

@Thorgal Aegirsson said.. "You only look at nutrition. "

Actually not at all true - I've written about exercise and training(they are not one and the same) here and other places before.

I remember seeing several obviously motivated people at the gym - showed up everyday - did long workouts on the treadmill for months - failing to lose weight. Inspired me to look up the research - there have been many studies looking for people to exercise and lose fat weight - with rather consistent failing outcomes. A good way to work up an appetite is to work out.

That being said, while exercise nor training has not turned out to be a magic bullet for people to lose weight, it is very important for health - particularly BG. Not all exercise is the same - weight lifting, 'training', quite rapidly changes fasting BG ( see Rippetoe etc) and helps reverse gracification of bones. Strength is a much better predictor of mortality than BMI etc.. Consumption of adequate protein is key to increase/maintain muscle mass(vegans beware).

Somehow, there is some animosity about low-carb vs low-fat vs starvation diets(hunger might make people grumpy?). Yet, I am not interested in the political side - I want to figure out what it is that is breaking. All the diets work temporarily and mostly fail long term. Yet, I tend to think that LC has a better track record - but more important, if someone has elevated BG - reducing it by reducing carbs seems a good plan. High BG kills, reduces quality of life. But eating LC to cover a metabolic damage is not getting at the root cause.

It only takes a 1% error in appetite to cause obesity - that's why I think slight changes in insulin sensitivity could be causative - as well as other things - but I don't know for certain. No one does. Thus the enigma. This slight error involves a complex control loop with many moving parts - insulin, leptin are important parts, but it could be something smaller that is causative. A small defect in the overall control loop is all it takes to change the weight set-point.

Normal, healthy children don't have trouble with high BG when they eat rice. I remember being in the Philippines in the 1980's and rice eaters did not get fat - yet today, with lard being replaced with cheap veg-oils, things are changing.

There is no need to eat those oils - seems prudent to avoid them until this all gets sorted out.

@normalcarb said.. "So low-carb may work by improving leptin sensitivity"

I've thought about that - fructose containing sugars seriously increase trygly - which some think blocks leptin sensitivity. There are studies that show fructose before a meal increases the average amount people eat. There is no need to eat sugar food(things like 'breakfast candy'(cereal)). Seems prudent to avoid refined sugar for many reasons.

The problem is people can stop eating sugar and don't go back to their normal weight - something gets broken...

Mark Calderwood, MD said...

Is there any research showing a cascading effect of insulin resistance, ie, liver first, then muscle, then adipocytes? Or are all cells equally affected by fatty acids? If so, would fat cells then be first in the cascade? Of course, the hypothalamic insulin resistance might be important in this cascade, though, I'm not sure what effect that would have on homeostasis (adipostat functioning). The reason why I ask this is that the chronicity of excess calories must play a role in "breaking" our homeostatic system. Or is it as simple as an adaptive response, such as in hibernating mammals.

karl said...

@Mark Calderwood, MD

Good questions

I've looked but not found studies simultaneously looking at insulin sensitivity in different tissues - (Seems an obvious need to have these studies - was asking petro about this a few weeks ago). Muscle tissue is different - training changes the sensitivity rapidly - yet the weight seems to stay. There are power-lifters squatting huge weights with huge muscle mass yet with excess abdominal fat.

Insulin effects a lot of things - in adipocytes LPL and HSL appear to be effected by insulin via a common system - but I don't think we know if their sensitivity tracks exactly. Either one out of kilter by 1% could cause problems. What happens if only HSL is overly sensitive?

One way to think of adipose tissue(many others can be seen as valid as well) is adipose is there to protect us from high BG. (High BG -> Liver pumps up DNL dumping trygly that adipose then puts into temporary(?) storage). ( High fructose is more dangerous than high glucose and REALLY turns up DNL to high gear - but not dependent on insulin from what I've read.)

The other sad bit is there are to many 'nutritional studies' that are done by people that don't seem to understand basic science. ( In real science(grade-school knowledge): we collect observational studies that we use inductively to form a hypotheses (these studies only show correlations - not causation) then we do real experiments with ONE variable to test the hypothesis. Changing from eating cornstarch to a mixture of different fats and sugar is not ONE variable. Using a control diet from one company and test diet from another - obviously different base feed lots - puts a lot of papers in the junk science pile. They used to do synthetic diet studies in the 1960's.

You said:
"The reason why I ask this is that the chronicity of excess calories must play a role in "breaking" our homeostatic system."

But do we really know that? Could it be that something breaks the system causing the excess intake - which breaks even more things? But what ever is breaking things - the damage sure appears permanent. Tell highly motivated people to restrict their caloric intake - and they will mostly fail.

I'm more suspicious of PUFA - - particularly linolenic acid - super high in vegetable oils - also high in (corn fed) chicken meat. It appears to mess with insulin sensitivity - directly causes weight gain. There is also conversion to AA(arachidonic acid) - really nasty stuff. Could it slowly break bodies - and then over-consumption?

The problem is when we look at people that HAVE ALREADY gained weight - we see BG and insulin sensitivity out of whack. Low insulin sensitivity is sort of expected if someone is overweight - and needed to release fats during sleep (morning hours). The assumption is insulin resistance is bad (we don't want high BG), but for adipose tissue in someone over-weight insulin resistance is good. What if what broke has nothing to do with BG directly? The dis-regulation could be just a consequence - or not. All we need is one little part of the whole system to break creating a fat transfer error of 1% or an appetite error of 1% and we have a problem.

Unknown said...

karl, you said "It only takes a 1% error in appetite to cause obesity"

This seems very wrong. If my healthy weight maintenance level requires 2000Cal, and I err daily by 20Cal, I maintain a new slightly higher weight in the long term. I certainly don't become obese if I err daily 1% over what a similarly built lean person consumes.

dbk_999 said...

karl, is your writing available online or elsewhere? just curious..

Unknown said...

My model for appetite feedback is a function of several variables, including: 1) current weight and healthiness (state variables), 2) palatability of food eaten (i.e., high-processed hyper-palatability, high-reward calorie dense foods, lower-reward whole foods; externally set by "diet" rules), 3) variety of foods available (cafeteria effect, also set by "diet" rules), etc. In other words, I see the feedback function as very adjustable by behaviours (setting rules, and kick-starting health).

My model accounts for "bi-modal" stability points of A) chronic obesity and addiction to fast food, and B) discovery of healthy eating and addiction to favourite exercise. I've been in both these modes, for many years each.

So far, the palatability model of obesity gives me the best insight into mastering my health.

Currently, I'm finding that the Potato Diet makes me spontaneously eat less:

Al Towers said...

Great post however, I disagree that elevated blood glucose and fatty acids disprove the "internal starvation" hypothesis. Is it not possible that these energy producing molecules are "trapped" in circulation, or constantly in flux in and out of tissues, without getting used for energy generation? I see the elevation of these in circulation to indeed be a form of tissue starvation, while the substrates are indeed available they just can't be used.

karl said...

@Kass Belaire
Not wrong - 2,000cal seems low for a healthy active person.. but I'll take the number anyway. Cutting intake of approximately 3500-calorie will lose about a LB - so 1% would be a lb every 175days - a couple of pounds a year - 10 years you are 20lb off target.

So if watching TV commercials or the size of your plate increases what you eat by 1% it matters. But there are all sorts of feedback that normally correct for it - until something goes wrong - thus my enigma: "what bit gets broken(tm)".

Also - losing that last 20LB is VERY difficult for a lot of people - why are the last lbs so hard? Is it that the set point gets changed? Why is it that abdominal fat is the last to go?

thhq said...

An advantage to using a CICO model is the flexibility to course-correct. 1% changes in eating are trivia compared to the 10% errors in the Harris Benedict equation, or the effects of sex/weight/age on metabolic rate. You can get a personal RMR if you want to go to the expense.

Or just walk a mile and raise your daily metabolic requirement by 4%.

Anonymous said...

CarbSane often cites the many problems of Gary Taube's Insulin model. There was even a recent post that mentions Robert Ludwig's new book.

Stephan Guyenet said...

Hi Mark,

The timeline of insulin resistance in different tissues is very difficult to determine in humans. However, we do have some evidence in rodents from my former colleagues. The vasculature develops IR before adipose, liver, and skeletal muscle.

We've seen an inflammatory signature consistent with insulin and leptin resistance in the hypothalamus in as little as 1 day on a fattening diet, supporting the possible involvement of this location early in the fattening process. This is the earliest detectable effect I'm aware of.

Hi Al,

I think that's a logical thought process. However, we have very strong evidence that disproves it. We know that people with obesity burn more calories than people without obesity, and that holds true for both carbohydrate and fat calories (as determined by measuring oxygen usage and CO2 output). So nothing is preventing heavier people from burning fat and carbohydrate-- they burn more of it than lean people.

That said, there is evidence that the brain can PERCEIVE an energy shortfall even when one is clearly not present. Basically, the brain measures a variety of signals of peripheral energy status to determine how to manage energy. But there is good evidence that it becomes resistant to these signals as weight increases, meaning that the brain thinks energy status is lower than it really is. So for example, when the brain becomes leptin resistant, it needs more leptin to feel like the body's energy levels are in the black, and more leptin means more body fat. So it's not true internal starvation, but you could say it's the perception of internal starvation by the brain, even in the face of plentiful energy and no barriers to fuel oxidation.

Unknown said...


You are making the error described here:

Even if you disregard the setpoint defenses (increased metabolic rate, reduced appetite), the new body mass comes with a metabolic cost. That's why there is a new equilibrium weight, not runaway obesity. And, that's a reason why runaway weight loss doesn't occur if you drop your intake by 100 per day. Reduced intake>reduced body mass>reduced caloric needs>new equilibrium weight.

karl said...


I'm on board with "brain can[might] PERCEIVE an energy shortfall even when one is clearly not present." Where the brain begins and the sensing tissue ends is part of the enigma - "what bit gets broken"..

Not surprised that obese burn more calories of all types - they have to - they have need more muscle to move their body - more to support digestions etc.

This gets back to my earlier points about What is hunger? What is satiety?

My subjective experience was that carb meals cause me to over-eat the next meal - not the carb meal. The next cycle of hunger was more intense. Others have made similar observations. (now how to measure this objectively?)

Anyone that has done any fasting knows that the initial hunger disappears - changes - there is still an urge to eat - but it is different. Why are carb meals in particular 'comforting' (my guess is serotonin ). How does leptin make people feel?

It would be possible to give people leptin, serotonin - mess with glucose levels, FFA etc-etc and do functional brain scans - say PET and see what lights up - correlate with questions about the subjective experience and more. Which signal or bit of nervous system is messed up?

At one time, it was expected that SSRIs would help people lose weight - but the post-synaptic receptors down regulate serotonin sensitivity.

RE The sited papers:

There are major problems drawing conclusions from those studies - In the second study the test diet and control diet are not from the same manufacture - so they are not even from the same feed lot - not possible to have a single variable. (having single variables matters!)

The high fat diet - does have lard in it (not all of them do) - has soybean oil (about 51% linoleic acid) and sugar Look at the differences in sugar..

So is the damage from the lard, sugar with linoleic acid? There are other studies that show that the fats are not damaging without elevated glucose. ( Sadly - best we have here are some correlations with multiple variables - with the misleading titles (why not "high fat AND sugar diets"?) it quits being real science. Even lard is not a defined substance - if we are ever going to figure this all out they need to do stuff like take a control diet add a chemically defined single types of fat - one at a time as a single variable. There was a time when research with synthetic diets in controlled situations was done on humans... I know - everyone does this sugar=fat bit - still it just isn't science... You can do two variable research - but it takes 4 groups not 2. )

Rodents in the wild don't get to eat much sugar - and are evolved for a really low "fat" (animal fat) diet. They do get a low amount of vegetable oils from seeds they eat. So how much stock to put in the results of a herbivore vs humans(omnivore)??

Real science is REALLY hard to do - 100 poor studies don't tell us what one good study can.

thhq said...

After a week of chewing a 500 mcg melatonin before each meal I observe the following:

There was no weight loss or gain. It's not weight loss magic.

My appetite was depressed mid-morning and mid-afternoon, resulting in much less snacking to stop hunger pangs. This would probably be a benefit if I was trying to go back to 500-1000 kcal/day weight loss deficits. But then again it might be no more than wishful thinking.

I stopped chewing one before dinner. The melatonin caused drowsiness in the early evening.

In summary a possibly useful trick for reducing hunger between meals but no reduction in total food consumption.

Anonymous said...

I would like to apologize for posting my question not regarding the issue being discussed but I was not able to get on the discussion for milk and cardiovascular disease. I am a heart patient myself and I enjoy a good glass of milk but have found out by researching that milk is the highest concern for the chemical dioxin. What I could gather was that dioxin is stored in the mammary glands of the cow, therefore having a very high level. I have been investigating as to what is healthier, grass fed milk, organic milk, non fat milk or no milk. May I ask for your opinion on this subject? I just bought organic grass fed milk today and was I am disappointed after what I have read. Thank you for any information you can share with me.

Teech said...

So fat people have the same concentration of fatty acids as thin people - shouldn't they be putting out more since they have way more fat? Seems like something is preventing all that extra fat they have from leaving their fat cells at a faster rate.

Will Kriski (Potato Strong) said...

The 'carbs raise insulin and insulin stores fat' theory forgets that you need fat to be stored in the first place, ie from the fat we eat. Glucose doesn't easily turn to fat in humans. Glycogen can be stored in the muscles and organs and burned off as heat. Amazing all the twists and turns we take to try to avoid the calories in and out reality. And at the end of the day, just look at Lustig to see if his theory works (ad hominem but very satisfying nonetheless).

Scientists love to complicate things but it's really simple. Lower the caloric density of what you eat (like all other non-domesticated animals) by eating plants, so that you naturally eat until full and don't exceed daily caloric requirements.

Unknown said...

Its 1% per day or per hour or per meal or whatever. So remember that the 1*1.01 equals 28% more after 25 cycles. Thats a lot. Plus it was said it was calories. Its 1% of the feedback loop. But its still 28% more after awhile.

TomW said...

There was an interesting report I heard recently about Cortisol related to coffee consumption. The gist being that Cortisol is at higher levels early in the morning until about 9:30 AM as the daily rhythm induces it to release glucose to kick your day off to an alert and effective start. The commentary suggested that drinking coffee between around 9:30-11:30 AM (and 1:30-5:30 PM) when Coritsol was lower would result in an improved coffee response. There was also suggestion that typical breakfast of sugar and sugar-like foods (bread, cereals, OJ) can lead to a sort of "Cortisol resistance". That is, if you amp the morning up with carbs and coffee your cortisol is not necessary for that function and so begins to lose effectiveness. A form of Adrenal atrophy?

Just another bend in the many complicated turns in the road to understanding nutrition and health. Sorry I can't post the reference and would also like to learn more about it.

Unknown said...

This isn't compound interest, it's simple interest. Overeating 1% is eating 1% over a healthy lean person's caloric maintenance level, not overeating 1% over what you ate the day before. Karl had that right, he just missed the new equilibrium weight.

Jim said...

Guyenet v Ludwig promises to exemplify a higher level of discourse than did Guyenet v Taubes. I thank you both for the exchange so far, and look forward to any future developments.

I'd love to see Ludwig address the possibility that a low carb diet's effectiveness (when it is effective) might be mediated by the fact that it tends to be a lower reward diet as well.

thhq said...

@jim @will The Carb Insulin assertion (it doesn't rise to being a hypothesis unless the proponents allow it to be rigorously tested) is a stalking horse for Atkins supporters. Significant lipogenesis only takes place in the absence of dietary fat. Overweight Americans do not lack for dietary fat. The way to coax them into the Atkins tent is to convince them that they are victims of low fat dietary choices. The fact that they are on average eating 50% fat diets is passed over by the sciencey books, blogs and articles in the NYT.

A stick of butter in your coffee is not a good substitute for a banana or a potato. Hang Atkins and all his wordy sycophants.

thhq said...

Having just read Taubes' "The soft science of dietary fat" from 2001, the figure of 34% dietary fat in the American diet is used. This number enters the urban mythology of HFLC and is often cited. I'd like to know where this number comes from if anyone knows.

The actual consumption of fat in the American diet is closer to 50%. 30-35% just in added fats, plus the fats from whole foods.

thhq said...

Here's another reference to increasing fat consumption, extended to 2010, which shows current American consumption at 50% of dietary calories.

As with the USDA's 1970-2005 dietary trends assessment, these results are the exact opposite of what Taubes asserted in 2001, and which the HFLC community uses autonomically to butress their low-fat-nation dogma. Total fat consumption increased from 34% in 1970 to 40% in 2000, to 50% today.

thhq said...

Colpo describes the source of the Taubes/Lustig decreasing dietary fat assertion as use of NHANES data

He also makes the case that the USDA increasing dietary fat trends (based on food disappearance) are much more accurate than NHANES (based on select group self-reporting).

Augh. Does this mean that the whole HFLC franchise hangs on using the right government data set? Are we high-fat-nation or low-fat-nation? The importance of carb insulin theory absolutely depends on low-fat-nation, as much as Ornish depends on high-fat-nation.

Thankfully my personal plans for health and longevity are not affected either way. Though I do like the USDA data set best.

thhq said...

Sorry to belabor this, but this thread's cooling off, and I'm on a roll....sort of...

It really doesn't come down to "correlation is not causation", or who ad hominem-ed who, or the fallacies of the Seven Nations Study. It comes down to One Nation with Two Studies. Neither is scientific, and both have major accuracy problems. The proper way to decide something as stupid as this is a coin flip. Heads: NHANES is right, we are low-fat-nation, sugar is the same as heroin, raisins are Satan's rabbit pellets, carbs drive insulin drives fat, and Ancel Keys is the devil. Tails: USDA is right, we are high-fat-nation, we eat nourishing pilafs made from grain grown in the shadow of the food pyramid, saturated fat will kill you stone dead if you even think about eating it, ketosis is a physiologically unsound diet (paraphrasing Ancel Keys from 1959), and Robert Atkins is the devil.

thhq said...

"Oh, no. No. No. I meant no scheme. I merely posed a little, academic accounting theory."

Leo Bloom, from "The Producers"

I'll pose my own little academic theory:

The NHANES 1965-2000 data showing the increase in dietary carbohydrate consumption is the basis for Taubes' carb insulin assertion. The Taubes scheme originated as a plausible explanation of the results of a highly questionable government survey. Science had nothing to do with it, and was added later for background scenery.,

And here's the corollary:

If correct, the USDA 1970-2005 data showing the increase in dietary fat consumption falsifies the carb insulin assertion. If % dietary carb consumption in the US did not increase, the carb insulin assertion explains nothing and it would be idiotic to propose it.

No wonder Taubes didn't mention USDA in his soft science piece. His scheme relies on soft statistics.

thhq said...

Mayo Clinic took a position on NHANES systemic error in using memory based dietary assessment last year, in line with the 2013 U South Carolina repudiation:

"Given the overwhelming evidence in support of our position, we conclude that M-BM data cannot be used to inform national dietary guidelines and that the continued funding of M-BMs constitutes an unscientific and major misuse of research resources."

Fifty years of flawed data collection down the toilet, along with the billions of dollars spent collecting it.

In contrast, Taubes is utterly dependent on NHANES for his increased carb consumption trend data. It is seen in his "soft science" article from 2001, and expanded on in Good Calories, Bad Calories on Page 232-234:

In fairness, Taubes didn't know about these flaws fifteen years ago, and used them in good faith assuming that they were accurate. Unfortunately he has used NHANES' flawed M-BM data to formulate his low carb dietary guidlines. It's time he retracted all the assertions he has made that depend on this database.

thhq said...

It's disappointing that the HFLC community has led, and continues to lead, the national health discussion off-point. While Eades/Taubes/Lustig/Wolf/Sisson were leading the charge by demonizing sugar, cursing Keys and the 7 Nations study, recommending eating more salt and putting butter sticks in coffee, and dismissing exercise as ineffective for weight loss; genuinely useful research was being done to provide more verification that the Med Diet is an effective CVD risk lowering technique:

HFLC thrives on parasitic deconstruction, and is always selling itself [as Colpo called Taubes' GCBC, an infomercial]. The more anti-conventional-wisdom the HFLC claim is, the better it is for the sales of the supplements and books. Testing whether the HFLC claims have merit, or retracting those claims when they are proven false, is off the table. Admitting that any other diet is a viable alternative to HFLC is also off the table.

Unknown said...

thhq said:

"Here's another reference to increasing fat consumption, extended to 2010, which shows current American consumption at 50% of dietary calories."


That graph shows added fat in food in pounds per person per year. That's not percentage of daily calories coming from dietary fat.

The graph shows about 50 lb of added dietary fat per person per year for the year 2010. That's about 560 calories per day per person (50/2.2*1000*9/365=560).

USDA data from 2008 show an average calorie intake per person per day of 2673. 560 calories of added fat amounts to about 21% of daily calorie intake (560/2673=0.21).

Unknown said...

USDA data from 2008 show 208 calories per day per person coming from vegetables and fruit. That's about 8% of daily calories. (208/2673=0.078)

Compare that with the 41% of daily calories coming from grains, added fat, and added sugar.

thhq said...

Coming up with apples-to-apples comparisons from USDA data has been difficult for me. I realize that the 2008 USDA report is in pounds, but it's useful for showing the continuing uptrend in fat consumption. I have believed for some time that average American consumption of dietary fat is around 50% of total calories. This came from reading the 1970-2005 USDA dietary trend report. Added fat in 2005 accounted for 32% of dietary calories, per the following:

"Added Fats and Oils. The Dietary Guidelines recommend that Americans keep total fat consumption between 20 and 35 percent of daily energy intake. In 2005, total added fats and oils available for consumption reached 86 pounds per person compared with 53 pounds per person in 1970. This 2005 estimate translates into 71.6 grams of added fats and oils per person per day after adjusting for plate waste and other losses. This estimate does not include dietary fats that occur naturally in foods, such as in dairy prod- ucts and meats. Added fats and oils account for about 32 percent of total calories for a 2,000-calorie-per-day diet. In short, the findings suggest that Americans, on average, need to cut back on added fats and oils because, while the 32-percent figure is within the Guidelines’ range, it includes only added fats and oils and excludes fats and oils naturally present in some foods."

We could argue about how much additional fat is naturally occurring - in meats, dairy, nuts and fatty vegetable matter - but it's probably in the 15-20% range.

I agree that we get a huge percentage of our calories as "added", both fats or carbs. I think that 41% is way too low for added calories. If added fats by themselves are 32%, added carbs would probably double that, and net/net, the average American gets 2/3 of calories from "added" fats and carbs. We're not a low-fat-nation by any means...more like drive-thru/jiffy mart/processed food nation. It's on that basis that I take issue with Taubes contention, in both GCBC and the NYT Soft Science article, that American fat consumption is decreasing. NHANES got it wrong and as a result Taubes got it wrong.

What's the story about Attia leaving NuSI?

thhq said...

This 2010 USDA report shows roughly 1500 calories per day from the added fats/sweeteners/grains, and 1000 calories from the whole foods (meat/veg/fruit/dairy). The 2500 total is up from 2000 in 1970, with slightly more calories increase in added fats than grains, and those two together accounting for 80% of the increase. Surprisingly to me the increase in sweeteners was only 34 calories, or 7%. [Based on the relatively small increase in sweeteners it's the chips that are causing the obesity, not the fructose-laden Big Gulps and raisins IMO].

Too bad we don't know what happened to our national metabolic rate CO. Even if CO stayed constant, the rising CI would result in a 20% increase in the national weight over the 40 year term.

annc said...

I am one of those who have lost a lot of weight and kept it off. Started at 250 and lost 75 pounds over two years, at three pounds a month, every month. For a decade I maintained it through careful but not obsessive monitoring, went up a few pounds during menopause but not significantly. Then I had a health crisis, took a very small dose of amitriptyline and within two weeks was starving, started gaining weight at a high rate, gained 12 pounds in less than two months. I finally got completely obsessive with cutting my calories down to about 1300 and, in spite of massive hunger, stopped gaining. But I did not lose a pound. A year later I went off the amitriptyline and thought my body would be more cooperative but in this last year, after many many sincerely consistent and obsessive efforts I have not lost a pound, I continue to have a very high level of hunger with even small cutbacks. I have tried every manipulation of carbs/protein/fats and overall calorie levels. Oh, and at that level of calories I have stopped sleeping--waking up at 1 or 2 or 3 and not getting back to sleep. If I move my calories up to over 2500 I can sleep, but I start gaining. Quite a dilemma. Whatever process is impacted by amitriptyline is significant for me, and it appears to not undo itself after the drug is stopped. Maybe my story can give all of you brilliant researchers some further clues to look into as to what impacts weight loss beyond the endless macro discussions and the total number of energy units consumed.

thhq said...

I found this NHANES study on the increase in American snacking from 1977-78 to 2007-2008. While this is based on the suspect food recall surveys, it gives an idea of the magnitude of the increase in snacking: from about 1 to 2 snacks a day on average, and a total snack consumption of 400 (F) - 600 (M) kcal per day. On average the addition of a 250 kcal snack per day.

Assuming NHANES to be approximately correct, this snack would account for 50% of the increase in calorie consumption USDA observed. This is equivalent to a 2 oz snack bag of Cheetos or Tim's Cascade Chips.

Regarding the increase in sweeteners, the 34 kcal per day increase is equal to 3 oz of Coke.

An extra Coke every 4 days + an extra bag of chips every day = obesity crisis

Unknown said...

thhq posted:

Added fat in 2005 accounted for 32% of dietary calories, per the following:

We could argue about how much additional fat is naturally occurring - in meats, dairy, nuts and fatty vegetable matter - but it's probably in the 15-20% range.


The USDA report by Wells that you posted stated the loss-adjusted per capita daily intake from added fat as 589 calories. The report chose to normalize to a 2000 calorie diet. This gives 32% (589/2000=0.32). It inflates the percentage because the actual average total calorie intake per person per day is significantly more than 2000.

Normalizing to the actual calories consumed would be more meaningful. The average per capita loss-adjusted daily calories intake in the U.S. in the 2006 to 2010 era was about 2600. This gives 23% of calorie intake from added fats and oils (588/2600=0.23). 23% is significantly lower than 32%.

The following USDA page gives data about loss-adjusted added fats and oils for 2006. It shows 639 calorie intake of added fats and oils and a total intake of 2685 calories. This gives 24% which is close to the previous figure of 23%.

Total fat intake is available on the CDC website. Mean total fat intake is stated as 33% of total calorie intake for adults age 20 and over. 33% is much less than 50%.

The CDC webpage states "data are based on dietary recall interviews of a sample of the civilian non-institutionalized population." This is different than the USDA method of using food balance data.

Trying to obtain total fat consumption from USDA balance data is not straightforward because its data is usually stated in terms of food groups (meat, dairy, added fat, etc), not macronutrient class (fat, protein, carbohydrate).

thhq said...

Paul D, if you want to believe that % fat consumption in the USA is decreasing that's your perogative. CDC's NHANES recall methods indicate that % fat consumption is decreasing. USDA's food depletion methods indicate that % fat consumption is increasing. But as the Mayo group pointed out, we cannot be simultaneously consuming more and less fat. Both methods are flawed. Maybe we should average them and say that % fat consumption is unchanged?

If % fat consumption is unchanged where does that leave the carb insulin assertion? Declining % fat supposed to explain the obesity crisis, but if % fat consumption did not change (or as USDA asserts, increased) the carb insulin assertion explains nothing about why we got fat, or soft science, or what is a bad calorie, or anything of any value in combating obesity.

Focus instead on the points where the USDA and CDC studies agree. By both accounts, total calorie consumption has increased by roughly 20% since 1970, and the majority of the increase is in added fats and processed grain carbs, and not in added sweeteners. IMO that is a better starting point for theories of obesity than in the supposed shifts in macronutrient %'s. Damning fructose and glorifying fats sells a lot of books, but it does nothing to get people to reduce their excessive snacking on salty/greasy/starchy Cheetos and potato chips.

thhq said...

After several years of listening to HFLC/Atkins/Paleo/vegan defenses, it was interesting to read about what these quasi-religious dietary systems have in common with actual religions. It starts with the narrative at the heart of the belief.

Whether it is the Garden of Eden, or the Cave of Paleos, or The Ring (Nibelung or Hobbit), or the Taubes/NHANES-34%-dietary-fat-and-decreasing, or the sacred-cow-whose-milk-is-only-for-baby-cows, the core mythology is unquestionable, and the True Believers are dedicated to defending it. They are the outer shields, protecting the core myth from any alteration by the infidels. They bring every dog to the fight, including apologetics, hermeneutics, cherry-picked scientific data, and ad-hominem attacks. They are doggedly persistent. Doing battle with them runs into thousands of words, until the infidel challenging the mythology stops arguing and gives them the last word. [And not because they are right, but because their game gets stale. Eventually you get tired of being called a liar because you don't recall correctly what Gandalf ate for breakfast.]

There is never any common ground with the infidel. Acknowledging that animal protein is healthy food, or that fructose is healthy food, or that dairy products are healthy foods, or that salty chips are the major cause of the obesity crisis, is not an option. In religious systems there is some accomodation to reality. Evolutionary adaptation cannot be denied, the Bible is not a science book, the Lord of the Rings and Wagner's Ring are works of fiction. The quasi-religious dietary systems are inflexible, persisting in their imaginary realities long past the point of credibility.

Jonty said...

Wow Peter, I'm so sorry to hear you've left NuSi! The promise of some answers to the questions posed in these comments kept me contributing, but now you're gone without so much of a hint of what questions may be answered, let alone when.

My 2 cents: The nutrition transition involves loss of health of traditional peoples as they transition to a Western lifestyle including sugar, refined fortified grains, vegetable oils, trans fats, xenoestrogens &c. Clearly, there's been a massive loss of micronutrients, too much iron and too little magnesium, a distortion of the w3/w6 ratio and the K/Na ratio &c. - and reversing these changes - from Weston Price's improving the dental health of children to the Lyon Heart Study - have unequivocally positive outcomes. Are we barking up the wrong tree pursuing macronutrient ratios? Should we not be looking at the quality of the diet? And shouldn't we be routinely measuring micronutrient status of diet study participants?

Unknown said...


You're putting words in my mouth and accusing me of having a prerogative. I didn't make any claims about total fat intake increasing or decreasing. I corrected you on a statement you made about a graph that you incorrectly said was indicating 50% of daily calories coming from added fat when the graph actually stated 50 lbs of added fat consumed per capita per year which amounts to about 21% of calories coming from added fat. You also appealed to a paragraph in a USDA report that normalized added fat intake to a 2000 calorie diet which gives an inflated figure of 32% of daily calories coming from added fat. Normalizing to actual total calories consumed (about 2600) gives a figure of about 23% which is consistent with other data.

These are issues of stating facts correctly, not ones of prerogative or religious beliefs.

thhq said...

@Paul, I've taken USDA annual nutrient availabilities from the following table, where macronutrients are expressed in grams per day. To your point, I'm trying to state the facts as correctly as the government statistics allow. [I hope the table I create below stays intact when this is posted]

Summarizing the results from 1970 and 2010, and converting to calories using 4 cal/gram protein or carb, and 9 cal/gram fat (I've deducted from the carb totals 19g dietary fiber in 1970 and 25g dietary fiber in 2010 as indigestible):

Calorie Basis
Year g protein g fat g carbohydrate cal protein cal fat cal carb % protein % fat % carb

1970 98 140 382 392 1260 1528 12 40 48
2010 120 190 449 480 1710 1796 12 43 45

Based on USDA availability % fat consumption rises from 40% to 43% of total calories. I stand corrected. This is less than the 50% total dietary fat I've claimed. From now on I'll use 43% total dietary fat.

This USDA comparison supports the Mayo group's finding. If USDA data is used, % fat is rising in the American diet. If NHANES data is used (see Taubes GCBC graph pg 232-4), % fat is falling. Both cannot be correct, and Mayo recommends rejecting NHANES.

One last point gleaned from this comparison. From 1970 to 2010, total available calories per day rose from 3180 to 3986, a 25% increase. While all macronutrients have increased, fat has increased by 36%, twice as much as carbs.

thhq said...

"33% is much less than 50%"

Forgive my presumptions but when I see 33% I think of the NHANES report that launched this statistic.

NHANES report of rising carbs and falling fats embedded 32.8% fat in the nutritional literature for many years. And not just Taubes. It survives to this day:

Here the blogger "consistent commentator" calls it up:

I could google it up a hundred more times. It was referenced in many nutrition texts in the 2005-2008 timeframe. It's a lie that won't die.

Please reference your sources @Paul D. Show me 33% fat without using NHANES. I admit that 50% is too high but 43% appears credible. Reviewing the USDA data it was 33% was in 1909 and has been rising ever since.

thhq said...

For what it's worth here's the report on Attia leaving NuSI

Makes it sound pretty shabby. No publications, no website updates, no press releases.

There was a journalist involved in this...what was that guy's name again?

thhq said...

I've just reread Stephan's responses to Dr. Ludwig again. I'm thinking right now about how the focus on carbohydrates has distracted from the central mission of obesity reduction. Through a number of diversionary tactics - fat won't make you fat, calories don't count, fructose is like heroin, salt we have misjudged you, for example - the focus on carbohydrates-only gives aid and comfort to the salty snack foods which are the major cause for the rise in obesity. Attacking sodas which contain 100% HFCS is laudable, but only amounts to 6% of the problem. 80% of the problem stems from salty foods containing 60% fat, 35% carbs and 5% protein (the nutrients of an empty Ruffles bag I found outside a grade school yesterday). Taubes, Ludwig and Lustig may decry foods like this too. But they're not doing it very loudly, since they contain lots of fat and no fructose.

JJ said...


JJ said...

Hi Stephan,

One of my undergrad students reads your blog and mentioned that our group got a mention. Thanks for the kind words. As an update, we have more published data (Templeman 2015 Diabetologia), and 2 big unpublished studies that will shed a lot more light on the subject, hopefully soon. Without giving too much away, in one study we find that acute reduction of insulin gene dosage after obesity can cause weight loss and in another study we find that a component of the insulin resistance that occurs with old age is caused by insulin (hyper)secretion itself (i.e. the desensitization concept). We have also tracked longevity in mice with modestly reduced insulin. Importantly, lipid homeostasis is more sensitive to modest (10-20%) changes in circulating insulin than glucose is. Anyway, we are learning lots of interesting things about insulin’s biological roles. However, as you astutely point out, it is critical to determine if every person’s beta-cell respond the same way to each of the macronutrients (or even fasting regimens). I’m not sure they do and we have (very) preliminary data to back up such an assertion. Finally, I don’t see how the 1st way, the 2nd way and the 3rd way have to be exclusive of each other. I think everyone knows that the brain is important, and indeed life sustaining, but certainly it can’t do everything and the pancreas and the hormone are there for a reason.

Stephan Guyenet said...

Hi Dr. Johnson,

Thanks for stopping by and sharing the update. I look forward to seeing your papers.

Unknown said...

thhq posted:

Summarizing the results from 1970 and 2010, and converting to calories using 4 cal/gram protein or carb, and 9 cal/gram fat (I've deducted from the carb totals 19g dietary fiber in 1970 and 25g dietary fiber in 2010 as indigestible):

Calorie Basis
Year g protein g fat g carbohydrate cal protein cal fat cal carb % protein % fat % carb

1970 98 140 382 392 1260 1528 12 40 48
2010 120 190 449 480 1710 1796 12 43 45


The USDA figures you are using are not corrected for losses to an extent that the data for daily calorie intake per person for 2010 is 3986 (480+1710+1796=3986). 3986 calories is inconsistent with measured body weight of the population and is about 50% higher than what people are actually consuming. This is one of the main issues of using food availability data derived from food balance sheets to estimate actual food consumption. Food surveys, questionnaires, and other methods try to avoid this by going more directly to the person or household and accounting for actual food consumption, although this too has significant shortcomings such as underestimation of calorie intake. Inaccuracies in food loss data and inaccurate personal recall are part of why USDA food availability data and NHANES survey data don't agree with each other in some respects such as NHANES showing 33% fat intake and USDA food availability data showing about 43%. Both seem to agree on significant increases in overall calorie intake since 1970.

When using food balance sheets, the losses of each macronutrient need to be accurately known in order to accurately determine the percentage of total daily calories consumed for each macronutrient. The figure of 43% of daily calories coming from fat may or may not be accurate. As much as a third of the fat in the food supply may be lost in the retail and consumer sector and as much as half may be lost in restaurants. Such losses have not been adequately considered in USDA food availability data. If unaccounted calorie losses of carbohydrate and protein are less than that of fat, then the unaccounted fat losses would inflate the calculated percentage of daily calorie intake from fat, and the actual percentage of consumed calories due to fat would be less than 43%.

Trends in the food availability data don't necessarily translate accurately into trends in actual consumption. As an example, a trend of a population relying increasingly on restaurant food implies that an increasing amount of waste of fat could be occurring, as restaurants may be wasting 50% or more of the fat in part due to deep frying compared with an expected lower waste at the household level. This would give the impression that there is a trend of increasing percentage of daily calorie intake occurring due to fat consumption when in fact it is an issue of higher unaccounted losses of fat due to increasing reliance on restaurants. As a hypothetical example, it could give the mistaken impression that the percentage of daily calorie intake from fat increased from 40% in 1970 to 43% in 2000.

Unknown said...

The following article makes reference to unaccounted losses of fat on page 25.

Nutrient Content of the U.S. Food Supply, 1909-2000
Center for Nutrition Policy and Promotion
Home Economics Research Report #56
November 2004


"While food supply estimates reflect trends in the availability of fats and oils for human food, they have never accurately measured the amount of food eaten because the portion of food wasted or discarded is difficult to determine. With the growth of the fast-food industry in the past three decades, it has become even more difficult to estimate the waste portion or discard of deep-frying fats. Since this discard is not available for human consumption, these estimates are limited as indicators of actual intake. A 1993 study estimated that about 50 percent or more of deep-frying fat used in food service operations is discarded after use and is not available for consumption (Hunter & Applewhite, 1993). Reliable estimates of total fats and oils are difficult to determine partly because the actual amount of frying fat discarded by food service operations, particularly fast-food restaurants, varies with the type of establishment. To better account for the actual use of fats and oils in the edible food supply, USDA’s Economic Research Service looked at food-loss factors at the retail, food-service, and consumer levels (Kantor et al., 1997). About one-third of the total fats and oils in the food supply were estimated to be lost through food service and consumer venues and thus not available for consumption. This study underscores the fact that food supply estimates for fat and oil are high; however, these losses are tentative and need additional research."

Unknown said...

thhq posted: What's the story about Attia leaving NuSI?


Regarding Peter Attia leaving NuSI, it would have been appropriate for NuSI to issue a timely press release commending Attia for his three years of service at NuSI and welcoming the incoming president. This is good public relations practice and general courtesy. Instead, everything was done quietly. People thought that the mostly dormant NuSI website was in error for not listing Attia as part of the group. Gary Taubes isn't known for courtesy and humbleness.

The following is a quote from a NuSI document.

"While scientific resolution is necessary, it alone is not sufficient; NuSI will also execute a sophisticated communications strategy to bring about societal change."

Nutrition Science Initiative
Introduction and Overview
September 2012

How is NuSI going to bring about a societal change with a sophisticated communications strategy when it didn't even have the courtesy to mention the departure of its president, Peter Attia, and left the public hanging?

The current profile of NuSI appears to be mainly venture capitalists and lacking in people with formal experience in nutrition and body weight management. The general public was hoping that NuSI would be more responsive to the needs of people who are interested in nutrition with regular updates of the NuSI website and timely press releases. A short publication of Attia's experiments using the metabolic chamber a couple years ago would have been useful. It would have at least been a publication by NuSI while the longer term experiments involving Ludwig and others are underway and whose results won't be published for a while. It may be an oversight on my part but I'm not aware of any formal publication of Attia's NuSI work in refereed science journals. If anyone knows of any, please post them.

Peter Attia has various interests and shows a desire for pragmatic treatment of health issues through various integrative approaches (exercise, meditation, diet), whereas someone like Taubes is more ideologically driven, corporate-libertarian, and fixated on blaming perceived incompetent health authorities and a particular macronutrient (carbohydrate) for the obesity epidemic. Taubes shows little interest and even a dismissive attitude toward modalities such as meditation and exercise. Such a difference in attitude can result in a schism that eventually causes colleagues to part ways.

Unknown said...

thhq said: "Show me 33% fat without using NHANES. I admit that 50% is too high but 43% appears credible."


Data collected and published by the USDA show energy intake due to dietary fat to be much less than 43% in the U.S. for adults for the 1994-1996 time span. It was about 33%. See the following USDA webpage and publication. USDA and DHHS were conducting their own surveys, CSFII and NHANES, respectively, before 2002. They combined in 2002.

Results from USDA’s 1994-96 Continuing Survey of Food Intakes by Individuals and 1994-96 Diet and Health Knowledge Survey Table Set 10

For percentages of total energy intake by macronutrient class, select "Report 10: Food And Nutrient Intakes & Dietary Knowledge 1994-96" from the webpage A pdf will download. Refer to Table 4 on page 13 of the pdf. For the 1994-1996 time span, the percentage of energy intake for U.S. adults was approximately

fat: 33%
protein: 16%
carbohydrate: 50%
alcohol: 1%

The data are broken down for male and female by age group. The figures I gave above are an average for the entire adult age groups (male and female).

For those who don't like survey data, the goal posts can be continually moved and any survey is not going to be acceptable, whether it was conducted by the USDA (CSFII), DHHS (NHANES), or any other group in the U.S. or other parts of the world.

Unknown said...

The paper in the following link has a table of the %E intake from fat for forty countries. Refer to Table 1. The data are generally of the survey type, as opposed to food availability estimates based on food balance sheets. Four countries in the table show a %E from fat greater than 40% (Cameroon, Greece, Russia, Slovenia). Eleven countries have a %E from fat less than 30%. The rest are between 30% and 40%.

South Korea and Japan have a %E intake from fat of 21.1% and 25.3%, respectively. They are not rice-eating peasants of China trying to avoid starvation that David Ludwig refers to in an attempt to dismiss inconsistencies in the carbohydrate hypothesis of obesity. In spite of being developed countries and having a relatively low-fat and high-carbohydrate intake, South Korea and Japan have rates of obesity that are low relative to other industrialized countries that have a higher %E from fat and lower %E from carbohydrate.

Stephan Guyenet said...

Hi Paul,

Sorry it took me a while to publish your comment about Peter Attia; it got stuck in my spam folder for some reason. I haven't spoken to Peter since this all went down and I'm not going to speculate about what happened, but I am curious.

Regarding fat intake around the globe-- good points. You can't just dismiss high-carb eating cultures as starving peasants.

ben said...

I beleive this is the link for Ludwig's response

Unknown said...

@ Ben Nguyen

That's the response of David Ludwig to the recent and soon to be published study by Kevin Hall.

There is heated debate on the blogs about the temporary 100 kcal increase in resting energy expenditure (REE) that occurred just after the large and abrupt macronutrient change in energy intake from fat from 35% to 80%. The increase in REE diminished toward its pre-jolt value after a couple weeks.

If I'm understanding the video of Hall at the conference correctly, there was a much higher activity energy expenditure of the test subjects outside of the metabolic chamber compared with inside the chamber, as much as 500 kcal per day per person. This happened even though test subjects in the chamber were exercising 90 minutes per day on an ergometer. The environment outside the chamber was a controlled indoor metabolic ward which attempted to keep activity energy expenditure similar to that inside the chamber, yet there was as much as a 500 kcal increase in energy expenditure. Even Hall didn't expect it. I assume that's a real energy difference and not a measurement fluke. It's a reason to not be dismissive of the possible effects of environment on physical activity. Casual observation may not notice or anticipate large differences in activity energy expenditure that could be occurring between two living environments that seem similar. Hall could devote an entire study to this 500 kcal anomaly.

The ongoing 500 kcal difference seems as relevant or more relevant to the obesity epidemic than the transient 100 kcal rise in REE. The general public is not likely to increase its energy intake from fat by 45%. Even if it did, it would accrue only a transient 100 kcal increase in REE according to the Hall study.

Fishermaaan said...

Good day Stephan. Best wishes with the launch of your new book - I will be buying this once it is published in the UK. I should also add that this very article is still my anchor point when I think about or refer to the predominant models of obesity. With that I have a question: you clearly identify the mind-gut axis (my words) as a key part of the problem. However, I don't see you refer to the microbiome. Given the huge strides in understanding in this area do you, even though you don't appear to have a specialism in microbiology, think (or allow) that this "organ" may play a role in the obesity riddle? Kind regards, Tom