Protein trumps carbohydrate and fat as the regulated parameter driving human food consumption:

Kevin Hall and Carson Chow published their estimation (in 2010) that the rise of obesity prevalence in the U.S. since the 1970s can be attributed to an increase in consumption of just seven calories per day per person. This is calculated by taking the average increase of energy stored in the heavier adult bodies of the present day, divided by the very long time frame it took for the change to occur.

The seven calories, on a daily basis, refers to the excess one would eat at that weight. Once we grow bigger, of course, the bigger body requires more calories to support it, so you would need a slow increase in food quantity over time to remain seven calories in excess daily.  When we look at the change in overall daily calorie consumption over the last 40 years (and this is the number we would need to target if we want to change obesity in the population) we now eat an average of 220 calories extra, per day, to maintain our bigger size (rounding off of NHANES data). So the change we are hoping to produce in the individual, depending upon where they are in their weight gain journey, would be somewhere between seven and 220 calories, daily.

This is somewhere between a sip or an entire 20 ounce bottle of Coca-Cola. The problem, and its solution, are both certainly too small to accurately track on a day to day basis by deliberately journaling one's food intake. The caloric understanding of human weight gain is limited by this fact.

To make progress, we must hope that there is something that we can do regarding what we are eating, rather than how much. Low carbohydrate diets seem moderately more effective than low fat diets, but neither solves the obesity problem, because they both produce fairly small amounts of weight loss and fail when participants go back to "eating as usual." If low carbohydrate diets show all the metabolic advantages we reviewed and work better than low fat or low calorie diets, why are they not a permanent solution?

One reason that we don't find any single diet strategy that works for everyone is that these strategies are targeted at individuals deliberately making conscious changes in their behavior. They depend upon each individual learning and implementing solutions to their own particular weight problem, as if there was something peculiarly wrong with their unique metabolism. However, the shifts that have occurred regarding our nutrient intake suggest that it is changes at the level of the entire food supply that should attract our attention for obesity causation and treatment.
In 2005, David Raubenheimer and Stephen Simpson quietly proposed a novel solution to the mystery of the obesity epidemic and coined the term "protein leverage" to re-frame the question of human obesity in terms reflecting animals interacting with their ecology to get the nutrition they need. I say “quietly” because, despite actively searching the obesity literature and attending conferences on the subject, I’d never heard the term until seven years after their first paper on the subject came out.

In a series of observations, experiments and sometimes abstruse articles, Raubenheimer, Simpson and colleagues have shown that many species, including humans, regulate food intake by how much protein is needed to maximize health and reproduction. The idea of “leverage” is used to explain the fact that small changes in protein availability can trigger large changes in animal behavior. When protein becomes less available, fruit flies will hold off on mating, humans will overeat and crickets will become cannibals.

I first became aware of their ideas in March of 2012, while trying to find academic articles studying “higher protein” diets as opposed to “low carb,” “Paleo,” “Zone,” or “Atkins” diets. I was looking for scientific validation of the diet scheme that I’d been advocating, which was to hold calories a bit lower than a patient’s typical eating, while increasing the proportion of protein in the diet. I came to this philosophy not by reading about it, but by reviewing the diet records of my patients. These almost uniformly showed that people with weight problems under-ingest protein compared to even the conservative USDA recommendations.

When we tried to decrease calories through simple portion control, the problem of low protein got worse, since smaller portions of low protein foods just reduced all nutrients indiscriminately. This, again was made clear by looking at the breakdown of nutrients in the food logs. To combat this, I began to recommend increasing the protein percentage. As I did that, I started seeing more positive results with my patients. The beneficial effect of the higher protein diets seemed to correlate with many popular approaches found in diet books. However, I’d never been able to swallow the strange Atkins assumption that carbohydrates are a “bad” food, or that we should all eat like cave men. Few naturally skinny humans need to purposely restrict carbohydrate to remain lean and nearly every animal eats some combination of fat, carbohydrate and protein and naturally gets to optimum health without excluding tasty food.

Something always seemed missing until I stumbled onto Simpson’s and Raubenheimer’s papers.

Unfortunately for the average U.S. layperson, these scientists work in ecology and zoology in Australia and New Zealand. So, while their ideas have appeared in recognized journals related to humans, such as Obesity Reviews, their theory has not caused much of a stir here. Because they are trying to bring together insights regarding many different species, the implications of their work are a bit buried in their papers in terminology and a graphical model they’ve invented called the “geometric framework."

In 2012, the two scientists finally put together the major findings regarding protein leverage into a beautiful, readable book titled The Nature of Nutrition-A Unifying Framework from Animal Adaptation to Human Obesity. While the geometric framework is still the centerpiece of their theory, the depth and breadth of the book bring even non-visual and non-mathematical readers to the understanding that protein may be the key to understanding the cause of our weight problems.

They demonstrate that most animals balance carbohydrate, fat and protein ratios within a fairly narrow space of optimal function. The animals either find themselves in an environment in which the natural balance of nutrients leads to good health without any regulation, or they regulate by choosing amongst a variety of imbalanced foods in order to get to the ideal. Through observation of how species act in the wild and by purposely changing the diet of study animals, the authors find, again and again, that species have an ability to detect what’s needed for optimum health and to generally feed in a way that maximizes fitness. When this can't be done, the individuals in a given species tend to adapt in a uniform manner, for example: sacrificing some reproductive health for living longer, or losing some size advantage to maintain energy. They call these adaptations the "rule of compromise" for that species.

Another possible means for overcoming poor nutrition in the local environment is to just pick up and move. When the environment changes (as when season changes causes protein to fall, or the food supply runs out) many insects and animals migrate. The most graphic example is the story of Mormon crickets, which form large migration clusters when protein runs low. The authors were able to show that the starving insects were driven to search not just for food, but high protein food, to the point that they would pass up many possible sources of energy (for instance, the grasses in the migration path) to get to an improved nutritional environment. So starved for protein were these insects, that if one of their brethren dropped dead from exhaustion and starvation, they preferred the fallen as a food source to the high carbohydrate fare along the way. 

The authors work their way up the food chain through salmon, rats and monkeys to finally explain what’s going on with humans. We, like many other species, seem particularly regulated by the amount of protein in the diet. While self-regulating human populations across the world and across several decades can be shown to eat a wide range of fat and carbohydrate calories, the amount of protein in the diet is nearly always 15% of the total. They compared the composition of the diets found in several countries to show that humans tolerate very wide ranges of carbohydrate and fat fluctuation.

Together, fat and carbs consistently add up to 85% of calories, but there is little pattern as to how much each supplies of that 85%. This suggests that carbohydrate and fat are interchangeable sources of “energy” while protein is the prime necessary nutrient the body will not go without. In a balanced, naturally regulated, human population with stable weights, the 15% protein number is found again and again.

So, what does protein leverage have to say about the “obesity epidemic" in industrialized countries? Simpson and Raubenheimer show that we are acting just like the crickets (well, not just like the crickets, thank goodness) and we are defending our protein target by a natural adjustment to a changed environment. USDA figures and another database called FAOSTAT show that the relative proportion of available protein in the food supply has decreased from 14-15% (just what we need) to 12.5%. This leaves us to feed in a sub-optimal environment in which the amount of protein we need to maintain health is harder to obtain. In fact, if we don’t consciously seek out protein and instead eat freely from the available food supply, we are in essence forced to over-consume fat and carbohydrate calories to reach our protein goal.

Because protein has been diluted by carbohydrate in the modern food supply, we must eat more carbohydrate to gain limited protein. Because food is so easy to obtain for most of us, we’ve made the switch to these lower protein foods at the expense of a couple  hundred extra calories a day without noticing much has changed... except of course, our weight, which we’ve sort been puzzled about, but not too puzzled to stop seeking the nutrients we need.

The insight of these scientists is in the concept of protein “leverage.” Historically, most researchers have discounted the importance of protein’s contribution to the obesity problem because it is such a small percentage of our total intake and because it is relatively constant across time and geography. Those two facts have made it very difficult to see that protein is actually the root cause of the obesity problem. Because we have to keep protein constant and because it is such a small component of overall intake, small changes in the percentage of protein have mathematical leverage to affect calorie consumption beyond expectations. In order to make up for a 1.5% decrease in the protein content in the diet, a medium sized person must consume 14% more total calories to reach the nutritional target.

Those calories currently come from an increase in carbohydrate, making it seem as if the carbohydrate itself has caused our obesity problem. But the protein leverage framework argues that the calories could just as easily come from fat, or a combination of both. Their role in this narrative is only as "nonprotein energy." Protein amount is small, but it can leverage large changes in the diet. The following example demonstrates the type of math found in their book:

A 150 pound man is at a stable weight eating 2400 calories per day with 14% of calories (336 cals) coming from protein.  The other 86% of calories (2064 cals) are made up of fat and carbohydrate.

The 336 calories of protein represent 84 grams needed daily, his protein target. If the amount of protein available decreases from 14% to 12.5%, he will have to make due with 75 grams per day. Or...he could begin to overeat the protein-dilute food to maintain the protein that his muscles and organs require for optimum function.

How many calories does he need to eat daily to maintain his 84 grams of protein?

84g = 336 protein calories. 336 is 14% of 2400.

If protein is reduced to 12.5%, what number is 336 12.5% of?

336 cals divided by 12.5% = 2,688 cals.

So if protein need is absolute and our bodies will technically be starving without our daily quantity, the internal regulation of appetite will unconsciously drive this man's consumption from 2400 to 2688 per day to stay healthy. Perhaps not coincidentally, the 288 calories in this example correlates fairly well to the amount of reported calorie increase in the American diet over the last 30 years, according to NHANES data.

Why would the body do this? Choose to be obese, or even diabetic, due to a change in the food environment? Because obesity is quite sustainable and doesn’t really decrease reproduction or fitness in a way that matters as much as protein starvation would. Diabetes, certainly, can be a horrible state to be in, but it’s preferable to letting the body remain low on protein, which would lead to poor functioning of muscles, organs, reproduction and longevity. Given the available options in the current food environment, the strategy we have adopted is to consume excess calories in order to gain the scarcer protein...and suffer the consequences.

Simpson and Raubenheimer call this the "Rule of Compromise" for human nutrition. We are obese because it is a reasonable trade-off, given our choices. Keep in mind, these are not conscious choices, but internal regulatory drives much too subtle, persistent and powerful to be over-ruled by our determination. You can diet for five or six months, but this system keeps track of what's best for you over an entire lifetime. It keeps track of what's best for our species as a whole.

Other theories of obesity attribute our over-consumption of calories to conscious or unconscious preference for sugar and fat. We are thought to be programmed by evolutionary pressure to take in calories when they are available, because famine is thought to have been common in our ancestors. By this thinking, some individuals and more particularly, some populations, have "thrifty genes" that make them want calories whenever they are available. The protein leverage hypothesis does not disprove the thrifty gene theory, it complements it. Or, perhaps it may make the thrifty gene theory unnecessary for questions regarding modern obesity.

If protein is the ultimate target for consumption, trumping our other needs, then carbohydrates or fat are being over-consumed, presently, not because they were once rare and our internal regulation treasures them, but because they are in the way of the protein we need. This allows us to propose solutions that do not require speculating on the dietary habits of hunter-gatherers.

Protein leverage, through a dispassionate comparison of humans with other animals, encourages us to think of obesity as an adaptation to the environment, rather than a disease state. In order to stay healthy, we must meet our protein target. To reach the target with our current food supply, we must overeat. To handle the excess calories, we store some as fat. Our ability to store fat is almost unlimited, so we grow indefinitely larger. With regard to handling energy, however, our adaptation is not unlimited. With regard increased fat and carbohydrate intake, for a time the body responds with a higher insulin level, which pushes blood sugar into muscles and other tissues. But, as this plays out over many years, our bodies become less sensitive to insulin's action.

Insulin's regulatory mechanisms, in a sense, being "overused" by constant presence of glucose in the diet, becomes less and less effective. Specifically, muscle and organ cells become unable to respond to the signal to open glucose channels to let sugar through the membrane. At that point, we are stuck with the carbohydrate, in the form of glucose, remaining in the blood or spilled in the urine as waste. Obesity and diabetes are simply the result of necessary over-consumption by the body seeking adequate nutrition. Perhaps neither problem would occur if we supplied enough necessary protein to the body in the first place.

This conclusion, while counter-intuitive at first, makes a number of other questions easier to answer:

Why do some obese people eat so much? Because they’re starving (for protein).
Why can’t we lose weight by eating less? Because that makes the starvation worse.
Why do we regain lost weight? The body is counteracting the worsened starvation.
Why does weight gain seem to accelerate as one grows larger? Because the heavier body needs even more protein, so must increase consumption faster.

When we put protein at the center of our focus, the futility of simply eating less, particularly “portion control,” becomes obvious: The body is actually more balanced, nutritionally, at the heavier weight, since it got to that weight by prioritizing protein. If you reduce how much you're eating, including protein, the body is wise to resist. It doesn't care as much about the number of pounds up or down as it does about the protein. Every time we reduce calories using a “balanced” approach, we reduce the number of protein grams even further and our body, naturally, compensates...just like an insect, invertebrate or primate living in the wild.

To accept this proposed explanation of human obesity requires a recognition that we share some of our primitive drives and the mechanisms for survival with the rest of the animal kingdom. The most humbling comparison is to consider the slime mold. This particular organism is literally a "blob" without organs, brain, nervous system, or even separate cells. Through very careful observation and experimentation with different nutrients, Dussutour and colleagues, in 2010, were able to show that slime molds slowly move toward particular nutrient targets over time. The "blob" was shown to have particular nutrition needs that it was able to satisfy by growing toward its preferred food over several days. Rather than be humbled by our similarity to slime molds, I simply think of it: "If blobs can do it, certainly we humans are capable of it too."

In determining the winner of the debate regarding obesity causation and the best diet for weight loss, I believe that Simpson and Raubenheimer's work allows us to finally settle the question in such a way that everyone can be happy: The low carb advocates are correct. The low fat advocates are correct. The low calorie advocates are correct. Even the high protein weight lifter types are correct. They are each correct due to the underlying principle of protein leverage.

You can either decrease fat, decrease carbs, or decrease both together in a low calorie approach, as long as you keep protein constant. Protecting the lean tissues of the body enables weight loss to proceed more easily. When low fat diets work, when low carb diets work, when low calorie diets work, it is because they maintain or raise the protein quantity in the diet. Simpson and Raubenheimer consider those diets to be special cases of the general rule of protein leverage.

This simple proposal, that protein in the diet regulates how much we need to eat, is different from the other proposals about obesity. The high fat hypothesis, the carbohydrate hypothesis and the "calorie is a calorie" hypothesis all deal with obesity as a disease. They seek the explanation for a particular problem that is occurring to a particular group of humans at this one time in our history. The protein leverage hypothesis is not just an explanation of obesity. It proposes to explain how our nutrition, in general, is regulated and how it relates to nutrition patterns observed in other species.

The authors are not actually proposing a new way of looking at obesity, they are asserting that they've discovered a "rule" of nutrition  (although they are careful to couch their proposal as a "framework"). If the proposal holds up to analysis over time, it would be, by my reasoning, really only the second rule of nutrition to be discovered. The first would be that energy into the organism must equal energy out (with allowances for losses through heat and waste). Moreover, if it is truly a rule, we should expect to see it manifesting wherever we look at diet.

In the paper by Austin and colleagues that we discussed in the carbohydrate hypothesis post, regarding NHANES data, the authors noted that protein and fat measured by grams remained constant while carbohydrate was the only macronutrient that quantitatively increased from 1970's to present. They don't speculate on the cause of the change and don't make reference to the idea of protein leverage. However, included in their discussion  of various analytic models they ran on the dataset, is a mention of a relationship very much in support of the Simpson/Raubenheimer thesis:

    "One of the most striking findings of this study was the consistently strong and negative association with increasing percentage calories from protein and daily energy intake across all 3 BMI categories...if protein was increased from 15% to 25% of energy intake in an obese individual, this would be expected to be associated with a decrease in energy intake of 438 calories (if substituted for carbohydrates) or 620 calories (if substituted for fat)."

Austin includes references to other studies which deal with high vs. low protein diets, but again, seem to have no knowledge of Simpson and Raubenheimer's work at the time of writing. The Nature of Nutrition was not published until a year later, so we can speculate (since Dr. Austin is not super-fast on answering emails from strangers) that these researchers discovered the rule of compromise in human nutrition independently, by scrutinizing their data and applying careful reason.

In 2015, a review of the effect of higher protein diets, published in the International Journal of Obesity by Arne Astrup, found evidence for protein advantage in adults who had lost weight and subsequently on their children. The European, multi-site study was designed to test diets of different protein quantity and carbohydrate quality on weight loss and normal daily intake of calories, in families. The increase of 5% protein in the diet resulted in significantly less energy intake and spontaneous weight loss in overweight and obese children who were not formally following a weight loss diet.

Dr. Astrup noted, "The most significant outcome of the DioGenes study was that very subtle changes in diet composition with respect to protein and carbohydrates seem to have a major impact on spontaneous caloric intake..." Again, there is no mention of Protein Leverage per se and no references to Simpson, Raubenheimer or their colleagues in the quoted references.

James Hill and his colleagues who follow the successful individuals in the National Weight Control Registry published an analysis in the journal Obesity in 2012, in which they sought to characterized diet styles, so that we may begin to customize diet advice more appropriately to certain types and dispositions. Previous papers from the national weight control registry could be used to argue that macronutrients don't matter for weight loss: The members of the registry don’t subscribe to any unified diet philosophy. Half of them use commercial weight loss programs, half don’t. Most exercise, but many don’t. Some have dieted many times, some were successful on the first attempt. It’s hard to look at these researchers’ papers and conclude that there is a particular method that is common to all successful weight loss participants.

However, in the 2012 paper, the authors included, in table 3, the macronutrient breakdown of the groups they were seeking to categorize. While the four subgroups differed in the percent of calories that come from carbohydrates (48-54%) and fat (26-31%), they all fell in a very narrow range for protein percent: 18, 18, 19, and 19%. This isn’t remarked upon in the discussion, it’s simply listed like everything else, but it is relevant to our current discussion. Across human populations, both in terms of geography and time, we find that protein is nearly always 14-15% of calories. Here we find a group of extremely successful weight loss registry members who consistently report eating 4-5% more protein than average in the typical human diet. While this finding is outside of the interest of the authors of the paper, it provides some useful numbers for testing our thesis.

Protein leverage would predict that if you shift a person from 15% of a 2600 calorie diet (averages from NHANES) up to 19%, then, in order keep grams of protein constant, the body would decrease calorie consumption by 410 cal/day, just to avoid protein excess. This alone might be the permanent compensation that has made 50-100 pound weight loss possible for these individuals. What makes this more interesting is that the participants were not specifically encouraged to eat more protein. They got to this balance by any number of methods. Could protein leverage be silently driving the success of these individuals while they believe they are purposely reducing calories, avoiding fat, "doing paleo" or any number of other methods?

The proponents of a low carbohydrate diet for obesity believe that switching from a belief system that blames fat for obesity to one that blames carbs for obesity would represent a major change in our thinking - a "paradigm shift." In a way, this is somewhat true, if we take it to mean that we will focus on how fat cells are regulated, rather than "calories in/calories out." But to me, fat vs. carbs is not a terribly exciting debate. We will run some more careful equal-calorie studies and pick the winner. A real change in thinking, worthy of the overused term "paradigm shift" (Dr. Kuhn would be appalled at what we've done to his concept), in obesity research would be to transition from behavioral, psychological and internal biological explanations to consider obesity an ecologic problem, as Simpson and Raubenheimer suggest. Perhaps with this wider scope, we may truly see where our nutrition has gone wrong.

Simpson, SJ and Raubenheimer, D. 2005. Obesity: The protein leverage hypothesis. Obes. Rev. 6:133-144.

Simpson and Raubenheimer, The Nature of Nutrition. Princeton University Press, 2012.

Dussutour, A. et. al. "Amoeboid organism solves complex nutritional challenges." PNAS USA. 2010.

Ogden LG, et al. "Cluster Analysis of the National Weight Control Registry to Identify Distinct Subgroups Successful at Maintaining Weight Loss." Obesity, Oct 2012.

Austin, Ogden, Hill, "Trends in carbohydrate, fat and protein intakes and association with energy intake in normal-weight, overweight and obese individuals." AJCN 2011.

Astrup, A. et. al. "The role of higher protein diets in weight control and obesity related co-morbidities." International Journal of Obesity. Jan 2015.