The short-term effect of a diet may have nothing to do with its long-term effect (2 of 2)

(Versión en español: pinchar aquí)

In the first part of this article we have seen an experiment that clearly shows that the CICO theory is wrong.

Let us assume that the following premises are true for a normal diet, not one that absurdly forces an excess of food:

  • In the short term, during the first week or the first two weeks after a dietary change, a low fat diet makes you lose more weight than a high-fat diet.
  • As time goes by, a physiologic adaptation happens and the roles of the diets are exchanged, being on average the low-carbohydrate diet better for body fat loss.

I am not saying that the premises are true, I only ask that we assume for now that they are true.

In this situation a person who I will call John decides to do a meta-analysis of weight loss studies and puts together in the same meta-analysis a) a dozen studies with a duration no longer than a week that mostly show a favorable effect for low-fat diets, and b) a few studies that are a little bit longer, a couple of months at most, which show a favorable effect for low-carbohydrate diets.

John mixes all the studies in the same meta-analysis and concludes that since no diet is clearly the best one, the composition of the diet is not that relevant and that what really matters are the calories! a conclusion that is actually in contradiction with each and every one of the individual studies. How do you feel? In this hypothetical situation that I am proposing, the composition of the diet would be key in the long term and the meta-analysis would have reached just the opposite conclusion, generating noise. A person who wanted or needed to lose weight and keep the reduced weight in the long-term would have to choose the right composition of the diet to achieve that goal.

Does anyone believe that the long-term effects of a diet can be inferred from experiments that are shorter than a week? Does anyone believe that the behavior of our body after months of losing weight has anything to do, anything at all, with what happens in the first three days of following that same diet? (see)

Obesity Energetics: Body Weight Regulation and the Effects of Diet Composition

I this article from 2017 its authors present a compilation of around twenty dietary studies. Table 2B shows us data on changes in body fat for these studies and concludes that, on average, low-fat diets help people to lose more body fat than low-carb diets do:

What are not shown in the previous graph are the durations of those studies. I copied the data from the ES column of the graph above (just as shown in that table, without checking the original articles) and I represented those values as a function of the amount of days the participants followed the diet. As we can see in the graph below, in half of the studies the diet was followed for one week or less. The duration of the study in days is represented on the horizontal axis.

Moreover, those studies with a longer duration, those where the diets are followed at least for a month, are favorable to low-carb diets (the one with the longest duration in the compilation did not use a low-carb diet, as I comment below, but two diets very high in carbohydrates):

The conclusions from the authors are amazing:

In other words, for all practical purposes “a calorie is a calorie” when it comes to body fat and energy expenditure differences between controlled isocaloric diets varying in the ratio of carbohydrate to fat.

Can you really deduce that from very short-term diet studies? It is enough for the believers in the energy balance pseudo-science, who, undoubtedly, use this type of articles to prop up their ideology, but for rest of us it is impossible to draw relevant, general or useful conclusions from this collection of experiments.

First, because of their duration: what is relevant is whether there are differences between diets in the long term, and in this compilation of studies no diet has been followed for more than two months. As a matter of fact, half of the experiments are no longer than a week. Do we want to know which diet is more effective in the long term? Let’s do the experiment, instead of making up a result from short-term data.

Secondly, the fact that some studies favor low-carbohydrate diets and some favor low-fat diets does not mean there are no differences between diets. At the beginning of this text I explained that if the differences were due to the duration of the experiment, by combining experiments of different durations in the same data pool the actual effect of the composition of the diet would be obscured in the average, when the reality would be that the composition of the diet would be key in the long-term effect of the diet. As I have said before in this blog, meta-analysis are another way of lying (see,see,see).

Thirdly, because all kinds of diets are being mixed in the comparison, from ketogenic diets maintained for a few days to diets that are simultaneously high in fat and carbohydrates that have absolutely nothing to do with healthy low-carbohydrate diets. For example, in the experiment from Rumpler et al. de 1991, the longest of all those considered in the compilation (see the last graph above), the high-fat diet was also very high in carbohydrates: 46% carbohydrates and 40% fat.

Can we infer from that result anything about a low-carbohydrate diet? Would the result have been the same if the diet had been ketogenic? The authors of the meta-analysis want us to believe that it would, but by including experiments like the one I am commenting in a meta-analysis, all they do is create misinformation.

Fourth, based on short-term studies, the authors of the meta-analysis reach conclusions that contradict the results of studies with longer durations (see). Are most of the long-term studies poorly done and their data is not reliable? Can we deduce that from 4-day long studies that have nothing to do with the long-term effects of the diets? Shall we ignore all the scientific evidence and replace it with the imagination/ideology of the authors of this meta-analysis?

Note, on the other hand, that not even the authors of the meta-analysis believe what they are doing. They downplay their own result by saying that a difference in fat accumulation of 16 g/d is “physiologically meaningless”.

Figure 2B shows differences in the rate of body fat change between diets with the pooled weighted mean difference of 16 g/d (P < .0001) greater body fat loss in favor of the lower fat diets (P < .0001). These results are in the opposite direction
to the predictions of the carbohydrate-insulin model, but the effect sizes are so small as to be physiologically meaningless.

But an energy imbalance equivalent to only 1 g d of dietary fat could explain the current obesity epidemic.

A small persistent average daily energy imbalance gap between intake and expenditure of about 30 kJ per day underlies the observed average weight gain (source)

Yes, this last statement comes too from one of the authors of the meta-analysis, Kevin Hall. He should explain why 16 g/d of difference between diets is “physiologically irrelevant”, as he says, but an imbalance of 1 g/d could explain the obesity epidemic, as he also says. They simply downplay their own result because it is so unbelievable, in the bad sense of the term, so erroneous, that it gives away that something is not right in its origin. Extrapolating this result to the long term makes it obvious that it is wrong. But, if it is not extrapolated to the long term, the authors of the article cannot conclude that “a calorie is a calorie”.

It is not the first time that Kevin Hall interprets very short-term results as a demonstration of long-term behavior (see).

What are the postulates of the energy balance pseudo-science?

We should notice that the energy balance pseudo-science is never explicitly and rigorously formulated in a way that its postulates could be falsified. Other theories are criticized and the followers of this pseudo-science argue that, as the other theories do not seem correct, “then a calorie is a calorie” (see). This is exactly what the authors of this meta-analysis do. It is typical of pseudo-sciences to avoid formulating their postulates so that they can be subjected to falsification. With the energy balance theory the absence of well-defined dogmas allows the coexistence within this pseudo-science of factions that defend postulates that are incoherent among them (see).

The consequences of all this charlatanism are very serious: public-health dietary recommendations are still based on the stupid energy balance pseudo-science, weight loss methods that have never been proved to work are still the official treatment for obesity and we continue to blame the victims for their failure to lose weight, arguing that they are not lean because they do not show enough adherence to the diet (see,see).

As a final note, the fact that something could only be accurately measured in specific conditions, does not mean that what we measure in those conditions is useful. Maybe only weight loss studies that last three days are really reliable, because you have the participants locked in a facility and you have absolute control about what they eat and what they do. You measure everything very well and you control everything very well, but the data that you measure is rubbish because the failure of the diets is a problem that happens after following the diet for several months (see).

Further reading:

The short-term effect of a diet may have nothing to do with its long-term effect (1 of 2)

(Versión en español: pinchar aquí)

One of the main dogmas of the energy balance pseudo-science is that when two diets have the same amount of calories and the same amount of protein, in that case they are equal for the control of our body weight (example). We are told that this idea derives from the First Law of Thermodynamics and that, therefore, to deny this dogma is to deny unbreakable laws of physics.

Let’s imagine that we do an experiment in which two groups of people are given much more food for a week than they would normally consume. Both groups receive the same amount of calories: one group receives 50% extra food in the form of carbohydrates and the other group 50% extra food in the form of fat. The same energy intake and the same percentage of protein. On the 7th day we measure how much body fat these two groups of people have gained that day. Should we get the same result from both dietary groups?

Is it possible, according to the energy balance pseudo-science, a result like the one I show in the graph below, where one of the diets produces more body fat accumulation than the other one?

No. It would not be possible according to that theory. This result would be in contradiction with the idea that our body weight is determined by the calories of the diet: the two dietary groups ingested the same amount of food in terms of calories!

How would the energy balance pseudo-science explain this result? It could not explain it and the reason is that that theory is nothing but charlatanism.

It is a real result, obtained from the following article.

Fat and carbohydrate overfeeding in humans: different effects on energy storage

For 14 days, 9 lean people and 7 obese people are given 50% more calories than the amount that is considered necessary for each participant. Each participant receives two types of extra food: one based on carbohydrates and one based on fat. The authors do not give details about the base diet nor about what the composition of the excess food is.

The evolution with time of the fat balance (difference between fat that is ingested and fat that is oxidized) is very interesting. Very interesting indeed.


As we can see, the result of this experiment shows that in those participants in the very first first days the “extra” dietary fat is much more fattening than the “extra” carbohydrates. But can we forecast, based on the previous figure, what will happen after day #14 (which is the day this experiment ends)?

It is impossible to ignore what we see in the figure above: not only the outcome is not determined by the calories of the diet —which is what the CICO theory postulates as obvious—in those participants (the result is a function of the composition of the diet), but we also found that it is irrelevant to know what happens in the first few days to know what will happen in the long term. We see what happens in the first 14 days of the experiment and we have no idea how the accumulation of fat would evolve from that moment on. We do not even know in what type of participants a diet can be more fattening than the other one in the long term.

The authors of the article apparently saw it differently:

we found that for equivalent amounts of excess energy, fat leads to more body fat accumulation than does carbohydrate.

Please note that they confirm that the CICO theory is dead.

But what I am most interested in is that this is a very short-term result, for all-male participants, for participants that are used to follow a high-carbohydrate diet and that are forced to eat a lot of extra food, extra food that is based on food products with a single macronutrient, not natural foods, etc. It seems to me that some people have serious problems limiting their conclusions to the conditions in which data have been obtained.

Do we extrapolate this result to people who follow a low-carbohydrate diet, who do not force themselves to consume more food than what their appetites demand, who do strength training, who follow a diet for years —instead of two weeks— and who consume real food, instead of half of their food in the form of a product that is 100% fat? Making that extrapolation is barbaric. In this article I want to talk about “scientists” who do that extrapolation.

This experiment is absolutely irrelevant for practical purposes, since it has nothing to do with the conditions in which a person would follow a diet high in fat and low in carbohydrates. Nobody defends a diet that is simultaneously high in carbohydrates and high in fat, such as the one that is used in this experiment. Moreover, in this experiment people are forced to eat in excess. But this experiment is useful a) as one evidence more of the falsity of the CICO theory and b) to demonstrate that short-term data are irrelevant for understanding long-term weight loss or gain.

The other major barrier to understanding is the focus on short-term studies. Obesity usually takes decades to fully develop. Yet we often rely on information about it from studies that are only of several weeks’ duration. If we study how rust develops, we would need to observe metal over a period of weeks to months, not hours. Obesity, similarly, is a long-term disease. Short-term studies may not be informative. Jason Fung

Further reading:

If your sugar intake today is lower than yesterday’s, do you slim down?


This article is an extension of an article that I posted a few days ago. The idea that I want to discuss is the one below:

We have data from the time evolution of two parameters, named A and B. Some people believe that there is a dependence relationship between A and B so that A has a significant effect on B, but we know that A suffered a trend change and B didn’t, so other people say that this fact suggests that it is highly unlikely that A has a significant effect on B.

For the purpose of explaining the failures in the previous idea, I will assume a very simple model of obesity: we gain 3 g of body weight for every 100 g of sugar consumed. Please, don’t bother to criticize this model: I will only use it as a tool to explain the errors in the idea explained above. What we will see is that, under the premise that sugar is determining our body weight gain, the correlation between sugar intake and body weight may be low. That is the main conclusion and the specific model used for the explanations is irrelevant.


As said above, our assumption is that we fatten 3 g for every 100 g of sugar consumed. If our intake of sugar were 50 g, we would fatten half that amount: 1.5 g.


Given a specific sugar intake, the DAILY increase in body weight will be directly related to that DAILY sugar intake.

If I’ve consumed a certain amount of sugar PER YEAR, the PER YEAR increase in body weight would also be directly related to the PER YEAR sugar intake. For example, if one year’s sugar intake were 33 kg, our body weight would increase by 1 kg that year. Had I consumed only one third of those 33 kg, my body weight would have increased that year one third of 1 kg.

Let’s say that over the years our DAILY sugar intake has changed according to the blue curve in the graph below. In this scenario, our DAILY body weight gain would change as indicated by the red curve (calculated according to the hypothesis of that we fatten 3g per each 100 g of sugar consumed).


DAILY sugar intake and DAILY body weight gain are directly related variables. Their correlation, i.e. the mathematically-computed resemblance between them, is maximum.

If we compute (in blue) the CUMULATIVE TOTAL sugar intake since 1980 (i.e. for each year we compute the total amount of sugar consumed since 1980 until that year), versus (in red) the CUMULATIVE TOTAL body weight increase since 1980 until that year, we get this:


Again, as we saw with DAILY body weight increase and DAILY sugar intake, there is a direct relationship between CUMULATIVE TOTAL sugar intake and CUMULATIVE TOTAL body weight gain. This is also to be expected: under the premise that body weight gain is directly proportional to the sugar intake, if the CUMULATIVE TOTAL sugar intake over the past X years gets bigger you are expected to gain more weight, and if it gets smaller you are expected to gain less weight.

In order to clarify what comes next, let’s assume we are filling a bucket by pouring into it daily glasses of water. Every day we pour into the bucket the contents of one glass of water. Let’s assume that the volume of water in the glass has been progressively rising, day after day, until reaching a peak at 110 ml, and then, for the last 15 days we have gradually reduced the volume of water in the glass until reaching 95 ml. Is the cumulative total water in the bucket expected to be reduced at the end of those 15 days? Does anyone think that if I reduce the volume of water poured daily, the volume of water in the bucket has to decrease? When we confirm it doesn’t decrease, do we conclude that it is highly unlikely that the volume of water in the glass has an important effect on the volume of water in the bucket?

Let’s get to the point, but first remember that the correlation between DAILY sugar intake and DAILY body weight gain is maximum, and remember, too, that the correlation between CUMULATIVE TOTAL sugar intake and CUMULATIVE TOTAL body weight gain is also maximum. Now, what kind of relationship exists between the DAILY sugar intake at a given year and the CUMULATIVE TOTAL body weight gain at that same year? A direct relationship is not to be expected: if the DAILY sugar intake is reduced, the CUMULATIVE TOTAL body weight is not expected to be reduced, in any case that year the body weight will go up by a smaller amount, but it will continue to increase, and the effect of a smaller DAILY sugar intake will also be small in relative terms, since we have only changed the data for one year that has to be accumulated to the rest of the years in the CUMULATIVE TOTAL: we have been accumulating body weight for years and the sugar intake during the last one of those years is expected to have a small effect in the CUMULATIVE TOTAL.

From another point of view, the CUMULATIVE TOTAL sugar intake —including the contribution from last year, but with a weight that depends on how many years are being considered—, is the variable which determines the CUMULATIVE TOTAL body weight gain to that year. It is nonsense to expect a direct relationship between DAILY sugar intake and CUMULATIVE TOTAL body weight gain. And, in fact, that relationship is not a direct one. Assuming that a man weighed 80 kg in 1980, the graph below shows his CUMULATIVE TOTAL body weight (red curve) versus his DAILY sugar intake (blue curve):


We know that the red curve is completely determined by the blue curve, but they don’t have a good correlation. Or in other words, we confirm that a low correlation tell us nothing about the existence of a dependence relationship between two variables. If for the same situation, we had chosen the two previous graphs, we would have concluded that the relationship between sugar intake and body weight gain is undeniable.

If we recall the idea at the beginning of the article, A (blue curve) has changed, and the effect on B (red curve) is apparently small, but we know for a fact that A completely determines B.

What are we seeing?

In the last graph I presented above, the red curve is completely determined by the blue curve and the correlation between them is low. How is that possible? Because these two variables, while they are completely related, they don’t have a direct relationship. One variable is the CUMULATIVE TOTAL body weight gain, the effect of several years of fattening, while the other variable is just the DAILY sugar intake for one of those years. The fact that one of them goes on rising, albeit more slowly, when the other one decreases DOES NOT suggest that a cause-effect relationship between them is unlikely. The CUMULATIVE TOTAL body weight gain is not supposed to go down when the DAILY sugar intake is reduced: a “negative consumption of sugar” would be needed to produce such an effect on body weight. And, in any case, it would be the level, i.e. the fact that sugar intake is negative, what would produce a decrease in the  CUMULATIVE TOTAL body weight gain, not the change, i.e. the fact that sugar intake decreases. Even if it were possible to consume negative amounts of sugar, neither there would be a direct relationship between both variables nor a high correlation would be expected.

anyone who defends that sugar intake is a main cause of obesity and diabetes is proposing that there is a direct relationship between those variables

That’s a fallacy. The hypothesis that sugar is fattening means that the DAILY sugar intake affects the DAILY body weight gain. Nobody says that the DAILY sugar intake is directly related to the CUMULATIVE TOTAL body weight gain to that date: it is stupid to expect that when your DAILY sugar intake goes down your CUMULATIVE TOTAL body weight has to go down. This is not mathematics: it is just common sense.

Does the data presented by Guyenet in his graph suggest, as he says, that “sugar is highly unlikely to be the primary cause of obesity“? No, it does NOT suggest that. His graph is absolutely consistent with a direct effect of the DAILY sugar intake on the DAILY body weight gain: it can easily be seen that when the DAILY sugar intake was decreased, the DAILY body weight gain also decreased.



The main Guyenet’s mistake, or the base of his attempt of deception, is that he assumes that the CUMULATIVE TOTAL body weight gain and the DAILY sugar intake are directly related, and that hypothesis is both nonsense and inconsistent with the hypothesis that he is trying to refute. No one proposes that the DAILY sugar intake has a direct relationship with the CUMULATIVE TOTAL body weight gain: this specific relationship is expected to be non-linear! Had he compared DAILY body weight gain with DAILY sugar intake, he would have found a direct relationship (consistent with the hypothesis that sugar is fattening). Had he compared CUMULATIVE TOTAL body weight gain with CUMULATIVE TOTAL sugar intake, he would have found a direct relationship (consistent with the hypothesis that sugar is fattening).

imagen_0430 imagen_0432

We have seen in this article, with help from a simple model of obesity, that although CUMULATIVE TOTAL body weight gain and DAILY sugar intake are not well correlated, that doesn’t suggest that there isn’t a causal relationship between both variables.

On the other hand, Guyenet confuses a lower intake of sugar with a negative consumption of sugar. Logic says that if sugar is fattening, reducing its consumption doesn’t make us slim down.


  1. When in this article I use the term “direct” what I’m saying is that when one variable goes up the other variable also goes up and that when one variable goes down the other variable also goes down. Note that a direct relationship is not necessarily one of proportionality.
  2. the data used by Guyenet as support for his hypothesis is epidemiological. This is relevant, since when part of the population chooses to decrease their sugar intake, they probably take additional measures related to improving their health, such as not eating grains/flour, not consuming processed products, cooking more at home, less frequently eating out, etc. And, in addition, those who make these decisions have not been chosen at random: they are the ones who have decided to take care of themselves, so we’re not just comparing sugar intake: we are comparing lifestyles. That is an important difference with respect to a randomized controlled trial (RCT), where participants can’t decide if they decrease their sugar intake (and perhaps its replacement by another product). In the case of a RCT that reduction is supposed to be the only difference between goups. Guyenet’s data is far from being that case.
  3. We are talking about total sugar intake, regardless of its format, regardless of when it is consumed, regardless of which products accompany it in the mouth. A lot of information is missed.
  4. We are talking about the average sugar intake of a population and the percentage of adults above a specific level of obesity. The interesting data would be to compare individualized DAILY sugar intake and DAILY body weight change, and we would want to have this data for a large number of people.
  5. Guyenet says the variation in the percentage of obese adult has been small —he even expected a decrease! — but, actually, the change has been bigger than expected from the small reduction in the DAILY sugar intake (gradually decreasing the intake from 110 g/d to 95 g/d is an average reduction of 9% relative to the baseline: it is not an 18% decrease!). May be people who have decreased their sugar intake have also taken, at the same time, other measures to improve their health, and those additional measures could be contributing to the undeniable trend change perceived from year 2000 in the percentage of obese adults.

Further reading: