A calorie is a calorie (more or less) after all

Just out in Cell Metabolism is Kevin Hall’s most recent paper that shows that low carb diets have no metabolic advantage over a low fat diet. In the experiment, a group of 19 individuals spent 22 days in total in a metabolic ward where their diet was completely specified and metabolic parameters were carefully measured. The individuals were put on both isocaloric carbohydrate reduced diets and fat reduced diets where the order of the diets was randomized over subjects. The short version of the result was that those on the fat reduced diets had more fat loss than the carbohydrate reduced diet although the cumulative difference was small. The body composition changes and metabolic parameters are also matched by the detailed NIDDK body weight model. You most certainly do not lose more fat on a low carb diet.

The results do show that a calorie is not exactly a calorie meaning that the macronutrient composition of the food you eat can matter although over long time periods the body weight model does show that macronutrient differences will always be small. Ultimately, if you want to lose fat, you should eat less and exercise more (in that order). It’s your choice in how you want to reduce your calories. If you like to go low carb then by all means do that. If you like low fat then do that too. You’ll lose weight and fat on both diets. The key is to stick to your diet.

This experimental result is in direct contradiction to the argument of low carb aficionados like Gary Taubes who claim that reducing carbs are particularly beneficial for losing weight and vice versa. Their reasoning is that carbs induce insulin, which suppresses lypolysis from fat cells. Hence, if you ate carbs all the time, your fat would get locked away in adipocytes forever and you would become very fat. However, the problem with this type of reasoning is that it doesn’t account for the fact that no one eats for 24 hours each day. Even the most ardent grazer must sleep at some point and during that time insulin will fall and fat can be released from fat cells. Thus, what you need to do is to account for the net flux of fat over the entire 24 hour cycle and possibly even longer since your body will also adapt to whatever your diet happens to be. When you do that it turns out that you will lose more fat if you reduce fat.

Now this was only for a diet of 6 days but experiments, funded by Gary Taubes’s organization, for longer time scales comparing the two diets have been completed and will be published in the near future. I’ll summarize the results when they come out. I can’t say what the preliminary results are except to remind you that the model has held up pretty well in past.

What is wrong with obesity research

This paper in Nature Communications 14-3-3ζ Coordinates Adipogenesis of Visceral Fat has garnered some attention in the popular press. It is also a perfect example of what is wrong with the way modern obesity research is conducted and reported. This paper finds a protein that regulates adipogenesis or fat cell production. I haven’t gone into details of the results but let’s just assume that it is correct. The problem is that the authors and the press then make the statement that this provides a possible drug target for obesity. Why is this a problem? Well consider the analogy with a car. The gas tank represents the adipocytes, – it is the store of energy. Now, you find a “gene” that shrinks the gas tank and then publish in Nature Automobiles and the press release states that that you have found a potential treatment for car obesity. If it is really true that the car (mouse) still takes in the same amount of petrol (food) as before, then where did this excess energy go? The laws of thermodynamics must still hold. The only possibilities are that your gas mileage went down (energy expenditure increased) or the energy is being stored in some other auxiliary gas tank (liver?). A confounding problem is that rodents have very high metabolic rates compared to humans. They must eat a significant fraction of their body weight each day just to stay alive. Deprive a mouse or rat of food for a few days and it will expire. The amount of energy going into fat storage per day is a small amount by comparison. It is difficult to measure food intake precisely enough to resolve whether or not two rats are eating the same thing and most molecular biology labs are not equipped to make these precise measurements nor understand that they are necessary. One rat needs to only eat more by a small amount to gain more weight. If two cars (mice) grow at different weights then the only two possible explanations is that they have different energy expenditures or they are eating different amounts. Targeting the gas tank (adipocytes) simply does not make sense as a treatment of obesity. It might be interesting from the point of view of understanding development or even cancer but not weight gain. I have argued in the past that if you find that you have too much gas in the car then the most logical thing to do is to put less gas in the car, not to drive faster so you burn up the gas. If you are really interested in understanding obesity, you should try to understand appetite and satiety because that has the highest leverage for affecting body weight.

Have we crossed peak food?

The New York Times has an article today describing the decrease in food consumption over the past decade.  Here is one primary reference. I used to joke that the obesity epidemic would eventually be curbed by either a huge increase in oil prices or a depression. The great recession of 2008 made be believe that food consumption would come down but the data shows that it may have been dropping earlier and mostly in families with children.  The biggest decrease is in sugar sweetened beverages.

Here’s Kevin’s mention:

The recent calorie reductions appear to be good news, but they, alone, will not be enough to reverse the obesity epidemic. A paper by Kevin Hall, a researcher at the National Institutes of Health, estimated that for Americans to return to the body weights of 1978 by 2020, an average adult would need to reduce calorie consumption by 220 calories a day. The recent reductions represent just a fraction of that change.

New paper on global obesity

We have a new paper out in the World Health Organization Bulletin looking at the association between an increase in food supply and average weight gain:

Stefanie Vandevijvere, Carson C Chow, Kevin D Hall, Elaine Umali & Boyd A Swinburn. Increased food energy supply as a major driver of the obesity epidemic: a global analysis, Bulletin of the WHO 2015;93:446–456.

This paper extends the analysis we did in our paper on the US food supply to the rest of the world. In the US paper, we showed that an increase in food supply more than explains the increase in average body weight over the duration of the obesity epidemic, as predicted by our experimentally validated body weight model. I had been hoping to do the analysis on the rest of the world and was very happy that my colleagues in Australia and New Zealand were able to collate the global data, which was not a simple undertaking.

What we found was almost completely consistent with the hypothesis that food is the main driver of obesity everywhere. In more than half of the countries (45/83), the increase in food supply more than explains the increase in weight. In other mostly less developed nations (11/83), an increase in food was associated with an increase in body weight although it was not sufficient to explain all of the weight gain. Five countries had a decrease in both food and body weight. Five countries had decreases in food supply and an increase in body weight and finally three countries (Iran, Rwanda, and South Africa) had an increase in food but a decrease in body weight.

Now by formal logic, only one of these observations is inconsistent with the food push hypothesis. Recall that if A implies B then the only logical conclusion you can draw is that not B implies not A. Hence, if we hypothesize that increased food causes increased obesity then that means if we see no obesity then that implies no increase in food. Thus only three countries defied our hypothesis and they were Iran, Rwanda, and South Africa where obtaining accurate data is difficult.

The five countries that had a decrease in food but an increase in body weight do not dispute our hypothesis. They just show that increased food is not necessary, which we know is true. Decreased activity could also lead to increased weight and it is possible that this played a role in these countries and the 11 others where food was not sufficient to explain all of the weight increase.

I was already pretty convinced that food was the main driver of the obesity epidemic and this result puts it to rest for me. This is the main reason that I don’t believe that the obesity epidemic is a health problem per se. It is a social and economic problem.

Journal Club

Here’s the paper I will be covering in Journal Club tomorrow:

Neurons for hunger and thirst transmit a negative-valence teaching signal


Homeostasis is a biological principle for regulation of essential physiological parameters within a set range. Behavioural responses due to deviation from homeostasis are critical for survival, but motivational processes engaged by physiological need states are incompletely understood. We examined motivational characteristics of two separate neuron populations that regulate energy and fluid homeostasis by using cell-type-specific activity manipulations in mice. We found that starvation-sensitive AGRP neurons exhibit properties consistent with a negative-valence teaching signal. Mice avoided activation of AGRP neurons, indicating that AGRP neuron activity has negative valence. AGRP neuron inhibition conditioned preference for flavours and places. Correspondingly, deep-brain calcium imaging revealed that AGRP neuron activity rapidly reduced in response to food-related cues. Complementary experiments activating thirst-promoting neurons also conditioned avoidance. Therefore, these need-sensing neurons condition preference for environmental cues associated with nutrient or water ingestion, which is learned through reduction of negative-valence signals during restoration of homeostasis.

New paper on steroid-regulated gene expression

I am extremely pleased that the third leg of our theory on steroid-regulated gene expression is finally published.

Theory of partial agonist activity of steroid hormones
Abstract: The different amounts of residual partial agonist activity (PAA) of antisteroids under assorted conditions have long been useful in clinical applications but remain largely unexplained. Not only does a given antagonist often afford unequal induction for multiple genes in the same cell but also the activity of the same antisteroid with the same gene changes with variations in concentration of numerous cofactors. Using glucocorticoid receptors as a model system, we have recently succeeded in constructing from first principles a theory that accurately describes how cofactors can modulate the ability of agonist steroids to regulate both gene induction and gene repression. We now extend this framework to the actions of antisteroids in gene induction. The theory shows why changes in PAA cannot be explained simply by differences in ligand affinity for receptor and requires action at a second step or site in the overall sequence of reactions. The theory also provides a method for locating the position of this second site, relative to a concentration limited step (CLS), which is a previously identified step in glucocorticoid-regulated transactivation that always occurs at the same position in the overall sequence of events of gene induction. Finally, the theory predicts that classes of antagonist ligands may be grouped on the basis of their maximal PAA with excess added cofactor and that the members of each class differ by how they act at the same step in the overall gene induction process. Thus, this theory now makes it possible to predict how different cofactors modulate antisteroid PAA, which should be invaluable in developing more selective antagonists.

Steroids are crucial hormones in the body, which are involved in development and homeostasis. They regulate gene expression by first binding to nuclear receptors that freely float in the cytosol. The receptor-steroid complex is activated somehow and transported to the nucleus, where it binds to a hormone response element and initiates transcription. Steroids can either induce or repress genes in a dose dependent way and the dose-response function is generally a linear-fractional function. In our work, we modeled the whole sequence of events as a complex-building biochemical reaction sequence and showed that a linear-fractional dose response could only arise under some specific but biophysically plausible conditions. See herehere, and here for more background.

Given the importance of steroids and hormones, several important drugs target these receptors. They include tamoxifen and raloxifene, and RU486. These drugs are partial agonists in that bind to nuclear receptors and either, block, reduce, or even increase gene expression. However, it was not really known how partial agonists or antagonists work. In this paper, we show that they work by altering the affinity of some reaction downstream of receptor-ligand binding and thus they can do this in a gene specific way. We show that the activity of a given partial agonist can be reversed by some other downstream transcription factor provided it act after this reaction. The theory also explains why receptor-ligand binding affinity has no affect on the partial agonist activity. The theory makes specific predictions on the mechanisms of partial agonists based on how the maximal activity and the EC50 of the dose response change as you add various transcription factors.

The big problem with these drugs is that nuclear receptors act all over the body and thus the possibility of side effects is high. I think our theory could be used as a guide for developing new drugs or combinations of drugs that can target specific genes and reduce side effects.