Tuesday, October 11, 2016

Did the US Dietary Guidelines Cause the Obesity Epidemic?

A popular argument holds that the US Dietary Guidelines caused our obesity epidemic by advising Americans to reduce fat intake.  Does the evidence support this idea, or is it simply a fantasy?


Low-fat diets have fallen out of style and are now a favorite punching bag in some corners of the popular media, the Internet, and the scientific literature.  And there are reasons for this.  As typically implemented, low-fat diets have largely failed to live up to their promise of reducing the risk of obesity, cardiovascular disease, and diabetes.  Randomized controlled trials over the last two decades have increasingly supported the position that higher-protein, lower-carbohydrate, and sometimes higher-fat diets promote greater weight loss than low-fat diets over the short- to medium- term.

The truth is that this is a significant medical reversal.  Researchers and public health authorities accepted a hypothesis-- and dispensed advice based on it-- to later learn that the hypothesis was not as well supported as they had thought.  This is fair to point out.

Yet this idea often gets pushed a step further: by shifting our diets away from fat and toward carbohydrate beginning in 1980, the US Dietary Guidelines actually contributed to the obesity epidemic.  This idea has been promoted in several popular books, and most recently, in an editorial published in the Journal of the American Medical Association (1).  Here is the relevant quote:
As a result of [US dietary guidelines], dietary fat decreased to near the recommended limit of 30% total energy. But contrary to prediction, total calorie intake increased substantially, the prevalence of obesity tripled, the incidence of type 2 diabetes increased many-fold, and the decades-long decrease in cardiovascular disease plateaued and may reverse, despite greater use of preventive drugs and surgical procedures. However, other changes in diet (such as meals away from home) and lifestyle (such as physical activity level) may have influenced these trends.
Recent research suggests that the focus on dietary fat reduction has directly contributed to this growing burden of chronic disease.
As usual, when someone makes a factual statement, they are stating a hypothesis.  Whether or not the hypothesis is correct depends on whether its key predictions are supported by compelling evidence.  So let's go through some key predictions of this hypothesis and see if the evidence supports them.

Prediction 1: There should be a correlation between the publication of the Dietary Guidelines and the obesity epidemic

The US government has been publishing dietary information for at least 120 years, and the 1980 version wasn't the first to recommend limiting the intake of fatty foods (2).  Still, it could be argued that the 1980 version was more strongly worded and more influential of public eating habits.

Americans have been growing fatter for over a century (3), but the trend has accelerated over the past 3-4 decades, and we call this acceleration the "obesity epidemic".  This epidemic was first clearly detected by the 1988-1994 NHANES survey.  The previous survey was conducted between 1976 and 1980, so we can say that it began sometime between 1980 and 1988 (4).

Roughly speaking, it is true that there's an association between the publication of the 1980 Guidelines and the obesity epidemic.  Yet by itself, this is very weak evidence, due to the large number of other things that also changed over that time period.  For the hypothesis to be credible, it will need more support.

Prediction 2: Americans should have reduced our fat intake in response to the Guidelines

If the Dietary Guidelines caused us to gain weight by putting us on a low-fat diet, there should be evidence that we actually began eating less fat in response to the Guidelines.  If not, the hypothesis cannot be correct.

Two independent lines of evidence suggest that our absolute fat intake did not decline after the publication of the Guidelines (5, 6).  Proponents of the hypothesis invariably cite the fact that the percentage of fat in the US diet declined, which is true (although the change was rather small).  The reason the percentage changed is not because our fat intake decreased, but because our carbohydrate intake increased, along with our total calorie intake.  Does this count as a low-fat diet?

As an analogy, imagine a man named Jim who has obesity.  Jim wants to lose weight, so he decides to eat a low-carbohydrate diet.  Rather than reducing his intake of carbohydrate, Jim adds fat to all his meals so that the percentage of carbohydrate in his diet decreases.  Jim's calorie intake increases from 3,000 to 4,000 Calories per day, and his absolute carbohydrate intake remains the same.  Yet the percentage of carbohydrate in his diet decreases from 45% to 34%.  Is Jim on a low-carbohydrate diet, and should we expect him to lose weight?

Of course not.  Jim isn't eating a low-carbohydrate diet, and neither have Americans been eating a low-fat diet.

3. Low-fat diets should cause weight gain

For this hypothesis to be correct, low-fat diets have to cause weight gain.  While low-fat diets as typically implemented may not be a very effective weight management tool, decades of research shows that they cause weight loss, not weight gain.  This is true even when we only consider studies that didn't involve deliberate calorie restriction (7, 8).

The fact that low-fat diets cause less weight loss than low-carbohydrate diets over the short- to medium-term doesn't mean low-fat diets cause weight gain.  Both diets are more slimming than eating as much as you want of everything.

4. People who followed the Guidelines should have gained more weight than people who didn't

If the Guidelines caused the obesity epidemic, then the people who actually followed the advice should have gained more weight than the people who didn't.  Yet the evidence shows precisely the opposite (9).  Here's the money quote:
Poor compliance with the Guidelines, rather than the guidelines themselves, is likely responsible for the weight gain observed in the American population. Adoption of an eating pattern consistent with the Dietary Guidelines should facilitate population weight control if sustained long term.
5. Other countries with dissimilar dietary guidelines should not have a similar obesity problem

I'm not going to do an in-depth analysis of how national dietary guidelines relate to changes in weight, but I'll simply point out that nearly all affluent nations have been gaining weight rapidly in recent years, regardless of dietary advice or lack thereof (10).  The US is only unique in the sense that we're on the leading edge of economic, dietary, lifestyle, weight, and health changes that are affecting the entire affluent world.

The verdict

Although the low-fat diet doesn't appear to be a silver bullet for our national weight problem, the hypothesis that US Dietary Guidelines caused the obesity epidemic by reducing our national fat intake has more holes in it than Swiss cheese on baguette.


Gretchen said...

I think one contributing factor is how the average person interpreted the Food Pyramid. Someone has pointed out how little carbohydrate you actually eat with your alloted 11 servings of CHO. But most people would look at that pyramid and think, "Oh I can eat all the white bread and white rice I want as long as I don't add fat."

Note I said "one contributing factor." There are many. And I think the hunger that, at least for some people including me, results from low-fat diets is another.

Austin said...

This was a timely post as I am currently reading through Marion Nestle's "Food Politics". It does appear that industry influenced recommendations in such a way that they are watered down. Between that and people's natural tendency to deflect blame, I think this makes the government guidelines an easy target for the media, diet books, etc.

Tucker Goodrich said...

"As usual, when someone makes a factual statement, they are stating a hypothesis."

Huh? Are you sure this is what you meant to say? If I say, "the sky is blue on a clear day", that's a factual statement, but not a hypothesis.

I think to say "While this appears to be a factual statement, it is really a hypothesis about the causes of these observed facts."

George Henderson said...

Are the 2005 Dietary Guidelines for Americans Associated With Reduced Risk of Type 2 Diabetes and Cardiometabolic Risk Factors?
Twenty-year findings from the CARDIA study

Overall, we found no association between DQI score and diabetes risk using Cox models adjusted for potential confounders. Higher DQI scores were associated with favorable changes in HDL cholesterol and blood pressure overall (P for trend <0.05), but with increased insulin resistance among blacks (P for trend <0.01).


However, you miss out a very important part of Ludwig's evidence - the 1,000s of low fat products created by government mandate, and the health-washing of these products by appeal to the dietary guidelines.

Also, as far as I know every comparable country with a similar obesity or diabetes epidemic did have dietary guidelines similar to the US, is exposed to US food products created in response to guidelines, or is exposed to WHO guidelines which were in essence the same as US. You underestimate the effect of these theories internationally if you think this isn't the case.

A further point, one I think is critical, is that the low-fat, low-animal fat hypothesis radically changed the dietary treatment of diabetes and obesity.
If this change had improved treatment, then incidence of these conditions ought not to have increased. Improved treatment of diseases typically results in population-level improvement.

Erik Arnesen said...

Another thing: The guidelines also recommended less sugar ("Avoid too much sugar") at least from 1980.

Sheila Kealey said...

Thanks for the evidence-based piece. Although many blame the Dietary Guidelines for the health problems of the nation, their sweeping statements aren't backed up with good evidence (too bad JAMA editors let this pass). And they often miss the critical piece that most people weren't following the guidelines. Also, the guidelines have never advocated consumption of refined carbohydrates, excess sugar, and excess calories, more likely contributors to health problems (even the first guidelines in 1980 recommended that carbs should be from vegetables, fruits, whole grains). http://www.sheilakealey.com/2016/04/19/fats-vs-carbs/

Though not perfect, the main themes of the guidelines have been relatively sound nutrition advice in line with evidence: including eating plenty of vegetables and fruit, choosing mostly unprocessed grains, reducing salt, fat and sugar, and getting more physical activity. The current guidelines improve by recognizing the importance of dietary patterns. Because blaming the Guidelines seems so common I felt compelled to write about it as well and will add your piece to the "more reading" section. http://www.sheilakealey.com/dietary-guidelines-really-making-us-fat-sick/

glib said...

Of course it did. It compromised the metabolic state of a substantial part of the population, and insulin resistance did the rest. Of course fructose is more damaging than glucose, and there are other factors, but for example when it comes to cars i thought that we started driving more and walking less long before 1980. The popularity of low carb diets derives precisely from large numbers of weakened metabolisms.

Ally Brown said...

Excellent comment by George Henderson.

I think it's perhaps better to ask a question along the lines of: Did the Dietary Guidelines help or hinder the development of an obesogenic food environment?

Mark Calderwood, MD said...

Of course, all forms of fat continue to be categorized under one roof. Subdividing into trans, poly, sat, monounsat would provide a more accurate picture of what really is going on.

John in Cape Charles Va said...

I'd say that food guidelines from the government are really beside the point ultimately. Individual exposure to food guidelines is at best minimal, for most perhaps not something that even mattered at all, but ultimately only marginally relevant to actual food choices. Food choice is quite a visual process and food cravings are complex amalgams of flavor, mouth feel, brain related drives and god knows what subliminal responses to stimuli. I was there in the fifties, sixties and seventies and I think the real reason for the obesity crisis is complex: sophisticated attractive packaging, store positioning of manufactured products, food and fast food restaurant advertising, the rise of color tv that elevated advertising impact, the rise of seed oils, artificial flavoring in processed foods, etc.

Fast food carry out restaurants had a race to the bottom in term of costs and therefore, quality. Fast food sit down restaurants followed and had a similar outcome. In both the goal is to maximize the emotional response at minimum cost.

We were victimized by capitalism, frankly. But no, food guidelines were not a part of the messaging in any meaningful way. I just don't see it as relevant.

Finally, I optimized my health on a low carb, medium protein, high healthy fat diet that also eliminated most restaurant food other than real food prepared on site. Perfect health at 66.

Gretchen said...

"the guidelines have never advocated consumption of refined carbohydrates, excess sugar, and excess calories, more likely contributors to health problems (even the first guidelines in 1980 recommended that carbs should be from vegetables, fruits, whole grains)."

Look at the picture illustrating the original Food Pyramid. The bottom layer recommends bread, cereal, rice, and pasta. It also shows crackers, and none of this stuff looks like whole grains. Looks like refined starches to me.

True, it didn't recommend excess food of any kind.

BJ Terry said...

I think this post shows a fairly un-nuanced view of cause and effect. Specifically, you are only considering the intended effects of the dietary guidelines, but not the potential for unintended effects.

For example, with respect to "Prediction 2" the intended effect of the dietary guidelines may have been that people would eat less dietary fat in absolute terms. But in the real world its possible that carbohydrates are less satiating than fat, and that the specific implementation of the dietary guidelines allowed companies to market highly palatable foods as healthy due to a lack of fat, both of which would cause people to eat more calories overall. This isn't the intended effect, but it is an effect, and the cause could be the dietary guidelines.

For "Prediction 4," (and 3 to some extent) part of the reason we have dietary guidelines is that most people don't have the resources of time or mental energy to deeply consider their food choices. Dietary guidelines are supposed to make it easier to think about how to eat healthily. Just because the people who DO have the time or mental energy to follow food guidelines lose weight doesn't mean that the guidelines haven't failed from a public health perspective.

Do I personally believe that the dietary guidelines are responsible for the obesity epidemic? No, mostly not. But I think these particular arguments against it are very weak, and do not necessarily follow from the quote you pulled.

The obesity epidemic is probably caused by advances in technology and wealth that have dramatically decreased the costs of calories and increased the palatability of foods. But I also think that there COULD be dietary guidelines that cause people to lose weight. For example, if the dietary guideline were "Only eat between the hours of 4pm and 10pm," its possible that the obesity epidemic would be less than it is now. On the other hand, maybe that would cause a traditionalist backlash that would cause people to eat more throughout the day out of spite, in which case the effect of this public health intervention would also be negative, despite it not cohering with the intended intervention! That's just how public health works.

thhq said...

The dietary guidelines are not in and of themselves a diet. In order for a diet to work it needs compliance. Also, the pyramid too vague to be of any practical use, because it uses imaginary foods. It would take a rocket scientist to figure out what category pizza and cheese nachos fit in, much less how much to eat.

thhq said...

"However, other changes in diet (such as meals away from home) and lifestyle (such as physical activity level) may have influenced these trends."

I would focus on this part before blaming the guidelines for the obesity crisis. My estimate using USDA macronutrient data is 43% fat in the 2010 diet, far from the 30% claimed in the editorial. And even 30% fat is a high fat diet by world standards. Americans are not eating, and never have eaten, a low fat diet.

The food guidelines themselves were completely ignored by the general public. The recommendation to reduce refined sugar consumption by 50% is case-in-point. If the guidelines had any effect at all, it would have been to create confusion for those few people who took them seriously. Many of the high calorie density obesogenic foods that are commonplace in the American diet are not shown on the Food Pyramid. Assuming that a person is trying to follow the guidelines, how is that person to decide whether the half pizza they just ate - composed of 50% fat, 50% carbohydrates, 50% meat, and 50% cheese (pizza is a manbearpig food, and defies normal math) - is a serving of grains, or a serving of fat, or a serving of meat, or a serving of dairy. And where does that take-out container of pad thai fit in, etc etc etc.

Kevin Bass said...

I'm confused. If fat calories decreased as proportion of calories, why does your graph at TEDxHarvardLaw show the opposite?


I have been referencing your TEDxHarvardLaw talk for some time, and I would like to know whether I should still be doing this, and what the reason for the confusion is.

Stephan Guyenet said...

Hi Tucker,

Any factual statement can be viewed as a hypothesis. Although it happens on an intuitive level, when you're looking up and determining whether or not the sky is blue, you're testing a hypothesis by making a prediction (the sky should look a certain way if it's blue), collecting data (visual image), and seeing whether or not your data support your prediction. If the data don't support the prediction (sky is gray), you don't say it's blue. You could also test this hypothesis more objectively by analyzing the spectrum of light coming from the sky.

Hi George,

I can believe that the availability and marketing of certain low-fat products was driven in part by the Guidelines, although much of this is a perversion of what the Guidelines recommended (e.g., eat less sugar). But are these foods more fattening than their full-fat counterparts (e.g., low-fat cookies vs. full-fat cookies)? I'm not convinced. Both versions are fattening, but I'm not aware of solid evidence that low-fat versions lead to higher habitual total calorie intake or fat gain.

Also, our absolute fat intake never went down, so it's hard to believe that Americans ever took the low-fat advice seriously. I do think the Guidelines had some impact, e.g. replacing whole milk with low-fat milk. But we made up for it in other ways, e.g. increased added fat and cheese intake. And other changes we made, such as increasing soda intake, can hardly be linked to the Guidelines, which specifically emphasized reducing sugar intake.

Hi Gretchen,

The pyramid image itself is not clear on whether the starches are refined or whole grain, although to me some look like whole grain, and others refined. However, the accompanying document clearly favors whole grains:

"Complex carbohydrate foods are better than simple carbohydrates in this regard. Simple carbohydrates such as sugars provide calories but little else in the way of nutrients. Complex carbohydrate foods-- such as beans, peas, nuts, seeds, fruits and vegetables, and whole grain breads, cereals, and products-- contain many essential nutrients in addition to calories.

Increasing your consumption of certain complex carbohydrates can also help increase dietary fiber. The average American diet is relatively low in fiber."

It's hard to find major fault with this.

Stephan Guyenet said...

Hi BJ Terry,

I agree with your general point that it's hard to be certain of the consequences when you start tinkering with a complex system. So yes, it's theoretically possible that there were unintended negative effects of the Guidelines. But this is also a hypothesis, and I'm not aware of convincing evidence to support it. So while it remains theoretically possible, there is no evidence basis to support a claim that it actually happened-- only speculation.

You hypothesized that carbohydrate may be less satiating than fat. This has been tested many times, and if anything carbohydrate is more satiating than fat, per unit calorie. It boils down to the fact that concentrated fat has a high calorie density and palatability. If you control for those two factors, carb and fat are equally sating per calorie. Protein is the only macronutrient that is different-- it's more satiating per calorie.

Hi Kevin,

Thanks for stopping by. There are two reasons for the discrepancy. The primary reason is that the time scales are quite different. In my TEDx talk, I was referring to the last century, a period over which our fat intake has increased, both in an absolute and percentage sense. In this post, I'm referring to the period from 1980 to the present.

The second reason is that there's an artifact in the USDA fat intake data that I must not have known about when I put that talk together. In 2000, the USDA changed how it gathers data on liquid oil intake, resulting in an artificial bump in fat intake. In my later work (and my book) I've corrected this artifact.

If you look at the graph of % macronutrients in my TEDx talk, you can see that the % fat in the US diet begins to decrease slightly, starting in the mid-80s. Then it increases abruptly in 2000 (the artifact). If you remove this artifact, % fat intake remains slightly lower than it was in 1980. But as I said, it's a very small effect.

guiseppe014 said...

With all the obsession with the circadian (24 hour) rhythm these days, it is surprising so little attention is devoted to the annual rhythm (and the monthly rhythm), especially in light of weight gain/loss. It must have been such that in the past people lost weight in the winter and early spring, and gained in the summer and fall. The adipose tissue seems to be where much vitamin D3 is stored, and so they would obtain vitamin D along with the fat they burn in the winter. The colder temperature may by itself have induced some additional fat burning to generate the needed heat, so perhaps one way to lose weight is to use less clothes in the winter and/or spend more time outdoors. What instead seems to be happening is that people eat even more in the winter (when even up until relatively recently it was not uncommon to eat very little the 40 days before Easter), and stay indoors in warmer temperatures due to lower heating costs, maybe watching more television or using computers than they did before. And maybe they try and lose weight in the summer, when they were supposed to gain weight etc. It would be interesting to see if/what changes have happened over the past century in these areas and if there is any correlation with weight.

Gretchen said...

Stephan, How many consumers actually read the Guidelines? I was in a clinical study of lipid meds at the Joslin Diabetes Institute and they just handed me a picture of the Food Pyramid and told me to follow that. I said I wouldn't, and they said, "That's OK. Diet doesn't make much difference anyway." This was in about 2001.

In the Food Pyramid era, "complex carbohydrate" meant polysaccharides as opposed to sugars. Only after they realized that some polysaccharides raise BG faster than sucrose did they change the definition to mean carbohydrate bound up with fiber that cause lower increase in BG, so they wouldn't have to admit they were originally wrong.

The difference between low-fat cookies and full-fat cookies is that consumers were deluged with ads from Snackwell's that suggested it was OK to eat lots of cookies as long as they were low-fat. I don't know if anyone ever studied the actual impact of this type of ad, but common sense says it would have worked.

Edward said...

In the 1980s, my wife worked as a biochemist at a pharma company (she later retrained to work in health care as an R.N.)

Her employer, as was typical at the time, held education sessions for staff and their families to educate us on the new USDA nutrition guidelines. I attended a presentation, with my wife, conducted by two certified nutritionists.

The gist of the presentation and Q&A can be summarized as: avoid fat, avoid as much fat in your diet as possible. Unless your are diabetic, sugar does not matter.

To emphasize that last point, (1) an audience question specifically asked about sugar and the nutritionist responded that except for tooth decay and diabetics, sugar consumption did not matter, and (2) I have a copy (now just a digital scan) of a USDA pamphlet from the era that said nearly the same thing.

That is the reality of actual professional nutrition advice promulgated in the 1980s and largely continuing in to the 1990s.

Mass nutrition advice of the era did not seem to be grounded in solid evidence, as history has shown, nor did it take into considerations that individuals are not an "average" and our bodies may process foods in ways different than "average". That I have food allergies (incorrectly diagnosed for 11 years) or my sister has celiac disease that was incorrectly diagnosed for 30 years are examples that recommendations must be tailored to individuals and not the masses.

Stephan Guyenet said...

I'd like to remind readers that the point I made in my post was about a specific mechanism that is often claimed: the Dietary Guidelines caused us to eat less fat and more carbohydrate, and that increase in carbohydrate caused us to gain fat. The post wasn't about the alternative possibility that the Guidelines caused obesity through some other unintended mechanism, although I'm skeptical of that as well.

Hi Gretchen,

I agree that's odd, but the USDA can't control what every dietitian thinks or says. You're right that complex carb used to mean polysaccharides, but in the 1980 Guidelines they define complex carbs as "beans, peas, nuts, seeds, fruits and vegetables, and whole grain breads, cereals, and products". This is clearly not referring to refined carbs.

Hi Edward,

I believe what you're saying, and yes that is probably bad advice, but again the USDA can't control what every dietitian says. The 1980 Guidelines are very clear in recommending a lower sugar intake and higher intake of unrefined carbs from "beans, peas, nuts, seeds, fruits and vegetables, and whole grain breads, cereals, and products".

I'm not aware of the pamphlet you're referring to, but I wouldn't be surprised by its existence. It's possible that there were mixed messages coming from the USDA-- I don't know enough about it to say. But regardless, the official 1980 Guidelines-- the ones that are blamed for the obesity epidemic-- were very clear in recommending a diet that's lower in sugar and higher in unrefined starch foods.

They do downplay the role of sugar in diabetes and cardiovascular disease, which frankly was not unreasonable at the time and is now uncertain but still scientifically defensible. But they acknowledge a role for sugar in obesity. In their advice on weight management, one of their four recommendations is to "eat less sugar and sweets".

George Henderson said...

The dietary guidelines may well have defined complex carbs by the real foods that supplied tham, and recommended lower sugar consumption.
The problem is that the people implementing the guidelines - in schools, in prisons, in consultation with industry, and in the media - didn't care about carbohydrate quality.
For example, in New Zealand we had the same guidelines (30% fat, 10% SFA, a trace amount of cholesterol) and - very recently - these spawned a "Health Star" rating system for processed food.
Products that are high in sugar and refined starch can get 5 stars (the highest rating) if they are high in protein, added vitamins, or added fibre (it is not at all necessary for this to be intrinsic fibre). Eggs have four stars presumably because of their cholesterol or fat content, while whole milk has fewer stars than skim milk or sweetened no-fat milk with added vitamins.

The people who created this system - experts employed by a government that has worked with USA-type dietary guidelines for decades - clearly still think that restricting fat, saturated fat, and salt are more important goals than restricting sugar or refined starch.

The dietary guidelines committee put a numerical limit on fat, saturated fat, and cholesterol.
They did not put numerical limits on refined carbohydrate, nor I think did they give any reasons why refined starch was to be avoided - they may have recommended some whole carbs (though not so much on the pyramid), but they did not condemn the refined.

Consequently the message that was most stridently pushed was "fat-free", "cholesterol-free", "low in saturated fat", and of course "heart healthy polyunsaturated oils" as the good fat.

This did create a lipophobia, which hindered effective early treatment of weight gain and type 2 diabetes. These are mostly reversible conditions, so their increase shows that treatment deteriorated. Treatment was strongly influenced by dietary guidelines.

RLL said...

The often maligned (and somewhat rightly so) Food Pyramid did recommend 6-11 carb servings a day plus fruit (call it 900 KCal). That is not necessarily excessive. But practically speaking most of us (me included) sadly deluded ourselves as to what a serving was. Other family members still assume grossly over size portions as a serving, but still think that avoiding fat makes it OK.

I think you let the Food Pyramid off a little too lightly.

thhq said...

Reviewing the USDA Food Availability data tables you cite, added fat consumption shifts up in 2000, but only adds to the steady upward trend. From 361 kcal/day in 1980 to 421 kcal/day in 1999, then a step change of 100 kcal/day in 2000 to 521 kcal/day, followed by a rise to 562 kcal/day in 2010. Subtracting the 2000 upshift, added fat consumption rises from 361 to 462 kcal/day, or +28%.

USDA cites the reason for the increase in obesity from 1970-2005 as a 500 kcal/day increase in caloric consumption, primarily from added fats and refined carbohydrates. Could the 1980 dietary guidelines have caused this increase in eating, even though the guidelines didn't recommend eating either of those calorie sources? I can think of two reasons why they could have. First the guidelines are not specific about kcals in a serving. The vagueness of 7-11 carbohydrate servings per day is unhelpful as well. As a result, the guidelines didn't put any hard restriction in place to prevent overeating. Second added fats and refined carbohydrates are usually eaten in a mixture, and the guidelines' encouragement to eat more carbohydrate servings brought more fat along as an unintended consequence.

Peter said...

Jason Fung says a lot of the increase in weight is due to the increased popularity of snacking, which keeps the hormones that cause weight gain going all day long.

Danielle said...

Increased snacking and multiple servings of carbs every day keeps insulin levels high. People permanently in the fed state will get fatter. It is not rocket science. Don't eat so often and it doesn't matter so much how many calories you eat. The guidelines recommended eating 6+ times a day so yes they contributed to the obesity epidemic. There is a fallacy that eating more often increases metabolic rate. Overweight people thought they were doing the right things and it made them fatter. Even normal weight people are more TOFI than they used to be just because they eat too often. Let's have a proper discussion about how long it takes for an obese person's insulin level to return to baseline when fasting, then let someone claim that these guidelines were a good idea for the obese, even if some normal weight people could kinda get away with it with regards to weight.

Stephan Guyenet said...

Hi Peter and Danielle,

The problem with this hypothesis is that overwhelming evidence suggests that elevated insulin is not a major driver of obesity.

garnetheartstrings said...


I know you are fascinated with debating the question of whether or not carbohydrates/fats causes obesity, as many of your posts are on that/deal with that question.

Your long-ago post on copper and heart disease fascinated me, and I have been looking into copper.

Please, what do you think of this news that copper is a fat burning mineral: http://newscenter.lbl.gov/2016/06/06/fat-burning-copper/


Haley Loughney said...

I feel like low-fat diets can be also considered a fad. People assume that because they are eating less fat they are getting less fat. I think that the dietary guidelines are accurate in portraying what an individual should be eating. My plate places emphasis on different portions of food an what must be included to hit all your nutritional categories. I'm not surprised that the results showed a lacking hypothesis. It's crazy that people following guidelines gain more weight than those who do not. I think people who pay more attention to food end up eating more.

Danielle said...

Thanks for your reply Stephan. Watching Dr Jason Fung's lectures on YouTube gives the impression it is the insulin doing the diabetes damage rather than the sugar. I believe he presented a graph showing that controlling blood sugar lower than 6.5 with meds resulted in no less death or CV events an so he suggested it was the elevated insulin doing the damage instead. I would really value you expanding on this idea and your debunking of Dr Fung's ideas as I am struggling to find this overwhelming evidence you referred to.

glib said...

It could be fatty liver instead, or poor gut biota, or nutrient deficiencies, or a combination of causes. We see a machine doing something (generating metabolic syndrome), the input (high density acellular CHO) and the output. Sure, it was not only the government, food companies could just sell cornmeal, potato starch and rolled oats and things would be better, instead they put fructose and trans fats everywhere. But the answer should be yes, the government guidelines had a hand in it.

Stephan Guyenet said...

Hi Danielle,

I'm not very familiar with Dr. Fung's work so I can't comment on it specifically. However, I will point out that the idea you proposed-- that the insulin itself is doing the damage in diabetes rather than elevated blood sugar-- is pretty hard to support. Before insulin therapy, uncontrolled diabetes was a death sentence, due to ketoacidosis and high blood sugar. Muscle and fat tissue atrophy. Insulin therapy restores metabolic control and allows a relatively normal life, although injecting it isn't as good as having a functional pancreas.

It is true that low-carbohydrate diets can help manage diabetes with less insulin, although the effectiveness vs. other therapies isn't as black and white as some low-carb diet advocates have proposed.

Regarding the role of insulin in appetite and obesity, I've written about it extensively on this blog. Here are a few posts:


I encourage you to consider why the carb-insulin hypothesis and related ideas don't get much traction among researchers (with the exception of a few vocal outliers), who are the people who know the most about the relevant physiology. It's not because there's a conspiracy.

Peter said...

I think the two parts of Dr. Fung's idea should be considered separately:

1. Does snacking cause weight gain and does not snacking result in gradual weight loss? I think so.

2. If so, is it because of insulin? I have no idea.

Thorgal Aegirsson said...


1. No, I can snack all day and still lose weight. Try to eat say an apple every hour, and nothing else. You'll see what happens ...

2. Re insulin: no. Unless you eat 24h / day and never reach the fasting state at night, no. Insulin is just there to shuffle nutrients where they should, and prevent concurrent fat circulation while alternative fuels are in circulation as well. As a side note, FFAs (free fatty acids) in circulation says nothing about fat loss. The rate of beta-oxidation over some time w.r.t to storage rate (net balance of fat trafficking) is what will dictate fat loss or gain. Non-diabetic people should stop focusing on insulin, it is a distraction at best.

karl said...

Fat is not one thing - carbs are not one thing. Confounding them as if they were single chemicals just muddies the water. (Ferinstance - fructose spikes trygly - while glucose does not).

What makes a much more interesting hypothesis - one that is testable - is when concentrated vegetable oils (PUFAs - mostly LA) entered the market mainstream in the early 1960's people just were not fat.

When I was first in the Philippines they still cooked in lard - only after they started marketing vegetable oils did I notice a increase in obesity. Because vegetable oils entered different countries at different times there should be a relationship to the increase in obesity in phase with the introduction and sales volumes.

Of course there is other evidence - if you want a fat lab rat you feed it sugar and PUFA( But it seems we are supposed to ignore that). So when people eat CIAB (Chips) they are eating carbs coated with PUFA).

glib said...

Speaking of Fung, his latest blog entry points out that that Kansas State professor who lived on Oreos for a month was funded by Coca Cola. Low carbers may be a fringe group, but using that result in public presentations (such as the AHS14 presentation) maybe requires a correction, no?

Stephan Guyenet said...

Hi glib,

Industry funding is common in nutrition research. If you look at the funding sources of leading low-carbohydrate diet researchers, and others frequently cited in the low-carbohydrate diet community, they also frequently have conflicts of interest. For example, Ronald Krauss (his work played an important part in Good Calories, Bad Calories) has received substantial funding from the dairy industry, and that fact has been used against him by people who don't like his work. Personally, I don't find his source of funding especially relevant since his work appears to be sound.

Industry funding is something that happens across the board, and it is good to keep in the back of your mind, but it doesn't automatically invalidate a person's work. I find it a little too convenient when people cite conflicts of interest to critique results they don't like, but overlook them when they like the results. I generally assume an academic researcher is doing their work in good faith unless I have a specific reason (something more convincing than the funding source) to think they aren't.

But in any case, Haub's "experiment" was really just an introductory story to get people's attention and get them thinking in the right direction. I like to open with something intriguing to get the audience engaged. My conclusions don't rely on Haub's experiment, and if you watched the talk, you saw that I presented much more compelling evidence of the importance of calories in later slides.

But Haub's results do make sense. They agree with a large body of evidence that a substantial calorie deficit causes fat loss and improves metabolic health in people who need it, regardless of diet composition. That doesn't make a junk food diet a good idea, but it does demonstrate the importance of calorie intake.

For what it's worth, I don't view the low-carb community as a "fringe group". It's well represented among the people who read my work, and although I like to challenge certain ideas that are common in that group, that doesn't change the fact that I think LC diets can be a useful tool, and for a subset of people, they work really well.

Patrick Rochon said...

problem is definition of Whole Grain. There's AHA stamps on cookie with ".. 7g of whole grains ..." or Cheerios with "8g of protein" (that actually comes from the milk that you would add anyway) things like that.


hence the Dietery Guideline encourage people to eat crap (empty calories) ultra refined (flour VS whole grain berries) etc ... all with AHA stamps of approval.

Same goes for margarine vs butter etc .... absolute amount of fat did not go down but much more omega-6 and less MUFA/omega-3, less cholesterol etc ...

Ken Adkins said...

Hi Stephan,

Link #5 appears to be broken. Can I ask what it is a link to?

Stephan Guyenet said...

Hi Ken,

Yes, it's a link to the USDA's food disappearance database. I don't know what's going on but maybe they're having a technical problem right now. I've been using that same link for years. They could also be upgrading and moving the data to a new location.

T33CH said...

Hey Stephen,

What about fat composition? I remember a post from you a long while ago where you looked at veggie oil vs animal fat consumption. You found that Americans did reduced the amount of animal fat they consumed (as recommended), and increased the amount of veggie oil consumption. From what I have read, high omega 6 seed oils + processed carbs is a big inflammatory no no.

GBE said...

Most ordinary people and scientists alike do not understand - or do not believe in it - the food reward mechanism and the addiction it causes.

It's a mechanism strongly related as those addicted to alcohol, cocaine, sex, gambling etc.

For example, we experiment a lot with rats - which i find totally inurat (inhuman) - 2 different groups are set up with available water and available water mixed with cocaine.
In one group young rats separated from their moms and one group full contact with their mom. A mom shows affection and support by licking her young rats and this group the young drink from water alone. In the group where the rats are separated from their mom, they drink the mix of water with cocaine. Why? because of the effect of dopamine and serotonin cocaine has on their brain, which those with their mother get naturally...

mydorazio said...

"The reason the percentage changed is not because our fat intake decreased, but because our carbohydrate intake increased...."

I am one of the poor slobs who took the dietary guidelines from 1980 to heart and I can tell you why that happened: by 9 pm I was starving! I shamefully downed 3 bowls of carb-packed bran flakes and could have kept going. The academics who stroked their careers with this hogwash should be reviled as among the worst perpetrators of suffering on mankind ever. I now happily eat bacon, eggs, and cheese to my heart's content and am never hungry and slim as a rail.

Jeff H said...

No mention in here of the affect on metabolic process and diabetes. Not sure this article is actually addressing the issue it claims to be addressing.

supervillainsomeday said...

Predictions 2 and 3 are not necessary to the hypothesis, unless your hypothesis is based on the notion that strict adherents to the dietary guidelines were the most likely to suffer from them. The hypothesis that the dietary guidelines caused the obesity epidemic can be approached more honestly with different predictions:

Prediction 2a: The guidelines caused a shift in manufactured/packaged foods and their marketing, in response to signals from customers that healthy foods were desired.

Prediction 3a: The low-fat trend in packaged foods led to an increase in overall carbohydrate percentage, as manufacturers substituted sugar for fat to maintain a good flavor profile.

Prediction 4: Overall consumption increased. (Secondary hypothesis: This was because low-fat foods are less satisfying.)

Prediction 5: Low-fat diets exhibited high rates of relapse, often resulting in rebounds above the dieters' initial weight. (Same secondary hypothesis as 4.)

The predictions you cited are built around the notion that as a rule, people actually pay attention to federal dietary guidelines when they eat. But that's fallacious, as often people make dietary choices impulsively. The dietary guidelines would have had a much stronger impact on food production and marketing, and therefore on impulse consumption--if you're acting on impulse, something marketing as healthy will seem better even if it isn't. For this reason I suspect my alternate predictions will bear out a heck of a lot better.

However, I do think it's unfair to blame those guidelines entirely for the change. You also have to consider that home cooking has taken a nosedive in terms of the percentage of meals we eat; this can be attributed to the rise of two-income families, the increasing demands of after-school activities, and the shift away from the nuclear family. With less home cooking comes more impulsive eating, and importantly also fewer nutritious veggies.

What we're looking at is a complex problem with multiple causes. But I think your predictions (except the first) miss the mark completely, and as a result you've grossly underestimated the impact of the guidelines by assuming--contrary to everything we know of human behavior--that their primary influence was a direct one.

Stephan Guyenet said...

Hi supervillainsomeday,

You stated that "The predictions you cited are built around the notion that as a rule, people actually pay attention to federal dietary guidelines when they eat. But that's fallacious". I agree that the Guidelines have little impact on our eating behavior-- it's a point I've made in the past. Yet my predictions don't depend on this assumption. I was responding to the specific claim, quoted in my article, that the Guidelines caused a reduction of fat intake and that this precipitated the obesity and diabetes epidemics. Pointing out that our fat intake never actually decreased is obviously relevant to that claim.

I don't understand why you are so confident that the Guidelines played a causal role in our increased refined carb/sugar intake. I have never seen any evidence of such a causal role, either direct or indirect via influence on the food industry. If you actually look at the 1980 Guidelines, you will see that they specifically and repeatedly advise limiting sugar intake, and in fact this receives just as much emphasis as the advice to reduce fat intake.

So why in the world would sugar intake have increased in response to this? Did the food industry see this clear and specific advice to limit sugar intake, and responded with an increase in the sugar content of its products?

Our changing diet probably had nothing to do with the Guidelines. It had to do with broader changes like the declining price of sweeteners (e.g., HFCS) and highly processed foods in general, replacement of home cooking with restaurant and prepared food, ubiquitous food availability and advertising, and advances in technology that gave us greater power over the sensory qualities of food.

It is plausible that the broader trend of eating less fat and not worrying much about sugar contributed to our diet shift, but the Guidelines are not part of that trend. And my suspicion is that the obesity epidemic would have happened with or without the increase in sugar/carb intake, because in the US our sugar/carb intake has been declining since 1999 yet obesity continues to surge. Right now our sugar intake is back to where it was in the late 80s.

Billy Gogesch said...

Thank you for making the effort to moderate this challenging topic.

I'm wondering out loud if our recent decline in sugar has not produced an effect in the obesity rate in the general population because it is non-linear. If there is a threshold above which our responses (on average) are saturated, then it may require a much larger decrease in sugar intake to see a response in obesity rates in the population.

Are you familiar with Robert Lustig's assertion that high fructose consumption is at the root of numerous metabolic disturbances which contribute to chronic disease. (summarized here: http://podcast.uctv.tv/webdocuments/Fructose-Epidemic.pdf) The prevalence of refined sugars in our diet, all of which contain fructose, has reached very high levels when compared to what was consumed in the early 19th century (mainly due to availability, I'm sure.) That is a very significant dietary shift caused by obvious preference coupled with the technology to make sugar available and affordable.

While he does not explicitly suggest the threshold effect I'm asking you about, the evidence in his thesis may support such an effect.

So, I guess I'm asking for your comment on the fructose thesis as presented by Lustig then the possibility that as a population we may be in a "saturated response" zone to sugars.

Thank you for indulging me in this speculation. I do think that this kind of speculative discussion is where valid research begins.

Kelly said...

I disagree with your second point because I believe many people went on a low fat diet without knowing it. When the notion that low fat foods were healthier became wide-spread many food companies began making "low fat" versions of their products. Often these new products completely replaced the old versions on store shelves and consumers ended up eating low fat food because it was often the only option in certain categories. For example, next time you are at the supermarket, try to find yogurt that is not marked as "light" or "low fat". Yogurt made from whole milk is virtually non-existent.

Grey said...


"I don't understand why you are so confident that the Guidelines played a causal role in our increased refined carb/sugar intake. I have never seen any evidence of such a causal role, either direct or indirect via influence on the food industry. If you actually look at the 1980 Guidelines, you will see that they specifically and repeatedly advise limiting sugar intake, and in fact this receives just as much emphasis as the advice to reduce fat intake."

This may be an age thing. I remember very clearly the switch to low-fat in *advertising* products as healthy. It was very dramatic and pretty much all the female members of my family followed that line ever since - with most of the men ignoring it as much as they were able. (They mostly didn't do the shopping so they got low-fat milk whether they liked it or not.)


"So why in the world would sugar intake have increased in response to this? Did the food industry see this clear and specific advice to limit sugar intake, and responded with an increase in the sugar content of its products?"

They tested low fat products in response to the guidelines and found people didn't like the taste - so they filled them with sugar instead.

So I think it's fairer to say the food industry followed one guideline and ignored another (because it was the only way to sell their products) - and nobody stopped them.

I take your word there's no evidence the increase in sugar was a *cause* of anything but I think the increase in sugar was clearly an *effect* of the low fat drive imo as I've seen them say so in documentaries.


(If iodine is mostly contained in oils/fat could the switch to sugar have lowered everyone's metabolism - so same calories but lower burn rate?)

Grey said...

"They tested low fat products in response to the guidelines and found people didn't like the taste - so they filled them with sugar instead."

I should clarify - there have been a few documentaries on this subject in the UK in the last year or so and on one of them a woman who'd worked in food taste testing since the 1970s said that's what happened.