Category Archives: Protein

Protein for Athletes

How much protein should athletes consume?

Bodybuilders have long known that consuming extra protein makes it easier to add muscle. Yet low protein dieting can enhance immunity against viruses and bacteria, and extends lifespan in animals.

The Perfect Health Diet, because we’re positive toward saturated fats and starches, will often lead to lower protein consumption than other Paleo diets that restrict fatty or starchy foods. So it’s natural that some athletes and bodybuilders have asked how to optimize protein intake.

Robert recently asked about this, but let’s look specifically at the case of Advocatus Avocado:

I believe my performance improved (albeit marginally–the differences aren’t large) when I allowed my protein/carb/fat ratios to remain consistent despite my high caloric intake, which is ~3,600 calories/day. In other words, I had a sense of better performance when I lowered my fat% to around 65 and allowed around 200g/day of protein (I work out 2-3x a week for an hour).

At 3,600 calories per day, 65% fat is 2340 calories; 200 g protein is 800 calories; that leaves 460 calories carbs. How do these compare with Perfect Health Diet recommendations for athletes?

Nitrogen Balance, Exhaustion of Benefits, and Toxicity

There are a few magic numbers for protein intake that we want to be aware of:

  • Nitrogen balance. Nitrogen comes into the body in dietary protein and leaves the body in urine as ammonia, urea, and uric acid after proteins are metabolized. So when a person is in nitrogen balance, the amount of dietary protein matches the amount of metabolized protein, and the protein content of the body is unchanged. Very likely, the muscle content is unchanged too.
  • Exhaustion of benefits. We want to find the “plateau region” for nutrients. Athletes want to know: at what level of protein intake does protein no longer help build muscle?
  • Toxicity. At what level of protein intake does protein begin to damage health?

Luckily Ned Kock of the superb Health Correlator blog has done much of the work for us in his post “How much protein does one need to be in nitrogen balance?.”

He presents this chart, from a book on Exercise Physiology [1]:

There’s a great deal of variability across persons. Some people are in nitrogen balance at protein intake of 0.9 g/kg/day; others need as much as 1.5 g/kg/day. At 1.2 g/kg/day, half the sample was in nitrogen balance.

Various factors influence the interpretation of this data:

  • The sample was of endurance athletes. Endurance exercise increases protein needs, so most people would reach nitrogen balance at lower protein intakes. Resistance exercise doesn’t require as much protein: Experienced bodybuilders are typically in nitrogen balance at 1.2 g/kg/day. [2]
  • Most of the sample probably ate a high-carb diet. Glucose needs were met from dietary carbohydrates. Low-carb dieters would need additional protein for glucose manufacture.
  • As Ned states, in caloric deficit, protein needs are increased; in caloric surplus, protein needs are decreased. If you’re restricting calories for weight loss, expect to need a bit more protein to avoid muscle loss.
  • Supplementing leucine “increased protein synthesis and decreased protein breakdown” [2], thus leading to nitrogen balance at lower protein intakes.
  • The point of nitrogen balance is dynamic: if everyone in the sample ate 0.9 g/kg/day, then they’d eventually get into nitrogen balance at 0.9 g/kg/day. The body adjusts to conserve muscle at given food availability.

The average person needs much less protein to be in nitrogen balance. The US RDA for protein, 0.8 g/kg/day, was set so that 97.5% of Americans would be in nitrogen balance. [2] But just to be conservative, and because we’re developing advice for athletes, let’s consider 1.5 g/kg/day as the protein intake that brings our athletes into nitrogen balance.

What about the protein intake that exhausts benefits?  At what intake is muscle synthesis no longer promoted?

Ned, citing a review paper [2], offers the following answer: “[P]rotein intake beyond 25 percent of what is necessary to achieve a nitrogen balance of zero would have no effect on muscle gain.”

On my reading it’s not so easy to infer a clear answer, but let’s go with this. If so, then muscle gains would be exhausted at 1.25*1.5 = 1.875 g/kg/day even for the most strenuous athletes.

What about toxicity?

We deal with this in our book (p 25). At a protein intake of 230 g/day (920 calories), the body’s ability to convert ammonia to urea is saturated. [3] This means the nitrogen from every additional gram of protein lingers in the body as ammonia, a toxin.

Clearly marginal dietary protein is toxic, via ammonia poisoning, at this intake level. A reasonable estimate for where toxicity begins is between 150 to 200 g/day.

Putting it together: A prescription for athletes

Let’s say our athlete is an 80 kg man. Then maximum muscle gain will be achieved at a protein intake of 1.875*80 = 150 g/day. Toxicity will begin somewhere between 150 to 200 g/day. So the “plateau region” where all the benefits, and none of the toxicity, are achieved is between 150 g/day and some protein intake not much above 150 g/day.

The plateau region is quite narrow! What this tells us is that athletes should consume about 150 g/day protein.

This assumes a high-carb diet, so that no protein is needed for gluconeogenesis. The body utilizes about 600 calories/day of glucose, plus another 100 calories per hour of intense training.

With carb intakes below 600 calories/day, additional dietary protein would be needed, because protein would be consumed nearly 1-for-1 with the missing carbs.

So we can summarize these results as follows:

  • On a high-carb diet (>600 calories/day), 600 protein calories/day maximizes muscle gain.
  • On a low-carb diet (<600 calories/day), 1200 carb+protein calories/day maximizes muscle gain.

Looking back at Advocatus Avocado’s personal experience, he eats a low-carb diet with 460 carb calories per day. We predict therefore that he would need 740 protein calories a day to maximize his muscle gain (plus up to another 100 calories per hour of training, to replace lost glycogen).

Advocatus says he needs 800 protein calories/day to maximize muscle gain. Close enough for blog work!

At these protein intake levels, Advocatus is probably experiencing mild ammonia toxicity. He might slightly improve his health by eating a few more carbs, and cutting his protein intake a bit.

He might also find that leucine supplementation would reduce his protein needs a bit.

Overall, however, I think his experiences are consistent with our framework for understanding nutritional needs. Those who are content with maintaining an ordinary person’s muscle mass can get by with relatively low protein intakes of 0.8 g/kg/day or less. But muscle-building athletes need high protein intakes, around 1.9 g/kg/day, to maximize the rate of muscle gain. If they eat low-carb, they may need even more protein. Such high protein intakes are likely to exceed the threshold of toxicity.


[1] Brooks, G.A., Fahey, T.D., & Baldwin, K.M. (2005). Exercise physiology: Human bioenergetics and its applications. Boston, MA: McGraw-Hill.

[2] Wilson, J., & Wilson, G.J. (2006). Contemporary issues in protein requirements and consumption for resistance trained athletes. Journal of the International Society of Sports Nutrition, 3(1), 7-27.

[3] Rudman D et al. Maximal rates of excretion and synthesis of urea in normal and cirrhotic subjects. J Clin Invest. 1973 Sep;52(9):2241-9.

Low-Protein Leanness, Melanesians, and Hara Hachi Bu

Gunther gatherer raised an interesting issue in a comment to Tuesday’s post.  Protein may be satiating in the short run; but what about the long run?

[H]igh protein keeps you full at first. But no one really knows for how long. Eventually it stops working and you find you’re eating a lot ON TOP of all the high protein you were already eating. All of us here tried Atkins long ago and fell off the bandwagon more than enough times to know it gets boring, stops working against hunger and doesn’t keep the fat off forever.

I agree. This is why we recommend a version of the normal Perfect Health Diet, which is normal — not high — in protein, for weight loss. Our diet isn’t the quickest way to lose weight, but we think it is likely to work best in the long run.

The Long-Term Effects of High-Protein Diets

I argued in Tuesday’s post that the satiating effects of protein had to be temporary, and that in the long run higher protein might cause, not reduced appetite, but only a slight change toward a leaner body composition.

A more interesting question is:  could high-protein diets be positively harmful?

Maybe!  Studies in both animals and humans indicate that eating high protein during childhood creates a predisposition for obesity in later life.

For instance, rats raised on a high-protein diet in childhood are more likely to become obese when given calorie-rich (sugar and fat) diets in adulthood. [1] The researchers conclude:

Our research demonstrating a significant susceptibility to an obese phenotype in rats weaned onto a high-protein diet and then challenged in adulthood with a high-fat high-sucrose diet suggests that lasting changes result from altering the composition of the first solid food that is consumed throughout growth into early adulthood. While all rats in this study consumed the same high energy diet during the last 6 weeks of the intervention, distinct metabolic profiles remained evident from exposure to the different diets during growth. This would suggest that these changes, either long-lasting or perhaps permanent, ultimately influenced the adiposity response of these rats to a high energy challenge in adulthood. Overall, it appears that a long-term diet high in protein, when mismatched with a high energy challenge, has negative effects on body mass and hormones and genes involved in glucose and lipid metabolism. [1]

The same phenomenon occurs in humans. In the book we mentioned a study showing that slightly higher protein level in infant formula – 9% protein vs 7% normally – caused children to become overweight two years later. [2]

So parents, let your kids follow their taste buds to high-carb, high-fat diets!

Might adult high protein diets promote later-life obesity too? I’m not aware of evidence, but I don’t think the possibility can be ruled out. I will look for evidence when I do research for two future blog series: one on protein intake, aging, and longevity; the other on connections between obesity and the human aging program.

Can You Be Lean on a Low-Protein Diet?

Tuesday’s post cited research indicating that we have a set point for protein intake: humans are genetically programmed to seek around 360 protein calories per day, and appetite becomes satiated once that is achieved.

But if protein intake determines appetite, then it seems those eating a low-protein diet face a Hobson’s choice:

  • If total calories are not increased, then the low protein dieter can expect to have a chronically unsatisfied appetite.
  • If total calories are increased, so that appetite is satisfied, then the low protein dieter can expect a higher equilibrium weight and a slightly less lean body.

Is it possible, then, to restrict protein, eat mostly carbs and fat which we know are the ingredients of appetizing desserts – and still achieve a lean healthy body, and feel comfortable? 

Yes, I believe so.

Melanesians Do It

Gunther pointed out that Melanesians are lean and long-lived on low protein diets:

I think it would be a bit fairer to include some consideration and explanation of Melanesians and their extremely low protein diet (anywhere from 10% to only 3% protein daily). Their extremely high level fitness and body composition flies in the face of all of these high protein studies.

It’s true: Kitavans, Tokelauans, and other Melanesians eat high-carb and low-protein, yet they’re noted for “extreme leanness.” [3] Why?

Well, first of all, the Melanesian islander diet is a variant of the Perfect Health Diet:  it is entirely free of food toxins. As I argued last week, eliminating toxins is the key to healthy weight regulation. But there are other factors.

Coconut Oil

The most abundant fatty acid in the diets of Kitavans and Tokelauans is lauric acid, the 12-carbon fatty acid which is the predominant fatty acid in coconut oil. [4, 5] These shorter-chain fatty acids are ketogenic and have significant effects on the body: for instance, they raise HDL levels. They also make people lean.

In one study, 8 weeks taking 1 tbsp per day coconut oil caused a significant decrease in body weight, waist size, and blood triglycerides. [6] In another, obese women who received 2 tbsp (30 ml) coconut oil per day slimmed their waist and increased HDL without an increase in LDL, while a comparison group receiving soybean oil did not slim their waist and had lower HDL with higher LDL and total cholesterol. [7]

Resistant Starch

Another feature of the Melanesian diet is that the biggest share of calories came from “safe starches” like yams, sweet potatoes, and sago.

These foods have a lot of fiber in the form of “resistant starch.” Digestion of resistant starch by colonic bacteria produces a lot of butyrate, a short-chain fatty acid that strongly promotes leanness.

For example, butyrate improves insulin sensitivity and prevents rats from becoming obese. [8]

Hara Hachi Bu

But the most reliable strategy seems to have been worked out by Asian and Pacific cultures long ago, many thousands of years ago.

The key is that if the diet is well nourishing, then appetite will be mild and easy to consciously control. What will be experienced is not hunger, which indicates malnourishment, but a mild desire for food that can easily be ignored.

There is an ancient Chinese saying:

“Eat until you are eight-tenths full, walk 100 steps after meals, live 99 years.”

In Japanese the saying is Hara Hachi Bu, eat until eight-tenths full. Hara Hachi Bu is common practice in Okinawa, where it helped produce the world’s most long-lived population.

As we note in the book, the traditional Okinawan diet is extremely close to the Perfect Health Diet: the Okinawan diet was rich in safe starches and animal fats, and near the low end of our recommended protein range: Okinawans ate about 300 g (2/3 pound) meat and fish per day.

The key here is that on a low-protein diet, eating until eight-tenths full is not a calorie-restricted diet. It is a calorie-sufficient diet that isn’t quite satiating because it is low in protein.

Intermittent fasting is a helpful part of Hara Hachi Bu. Eating only within a relatively short window each day makes it easy to keep calories down.

My Experience

I have been practicing protein restriction for several years, eating coconut oil and safe starches for several years, and practicing daily intermittent fasting for over six months. I drink tea or lemon-flavored water through the morning, and on most days eat only between 2 pm and 8 pm. If I snack during the fast it is usually either a piece of dark chocolate or some coconut oil.

I sometimes go many days eating relatively little, then eat a lot for a few days. Physical activity – sports, intense exercise, running – increases my appetite noticeably. Regardless, I never feel hungry. It is easy to complete the daily fast: some tea, sometimes with coconut oil or dark chocolate, is enough. I now do total fasting on religious fast days – for instance, I now do a 64-hour fast from Holy Thursday through Easter Sunday morning. Generally, as soon as I focus my attention on work I forget that I wanted food.

(This is a very good sign for my health, by the way. During my long chronic illness I couldn’t tolerate fasting at all.)

I haven’t noticed difficulty adding muscle when I work out. I believe I would add muscle more easily on a higher-protein diet, but as it is I add muscle more easily than I used to on the standard American diet. I’m content with my body composition.


I am convinced that by eating coconut oil, getting carbs from safe starches, and practicing intermittent fasting and Hara Hachi Bu, most people can become well-muscled and lean on a low-protein diet without any sense of hardship.

Since a carb-fat mix is the classic recipe for dessert, this makes for an extremely tasty diet.

If you want the surest way to lose weight over the next few days, eat a high-protein diet. But if you want a long-term diet that maximizes longevity by restricting protein, consider eating a tablespoon or two of coconut oil per day, fasting for 16 hours a day, and finishing your meal eight-tenths full.  

If you walk a hundred steps after dinner, you just might become a healthy centenarian!


[1] Maurer AD et al. Consumption of diets high in prebiotic fiber or protein during growth influences the response to a high fat and sucrose diet in adulthood in rats. Nutr Metab (Lond). 2010 Sep 29;7:77.

[2] European Childhood Obesity Trial Study Group. Lower protein in infant formula is associated with lower weight up to age 2 y: a randomized clinical trial. Am J Clin Nutr. 2009 Jun;89(6):1836-45.

[3] Lindeberg S et al. Haemostatic variables in Pacific Islanders apparently free from stroke and ischaemic heart disease–the Kitava Study. Thromb Haemost. 1997 Jan;77(1):94-8.

[4] Lindeberg S et al. Lipoprotein composition and serum cholesterol ester fatty acids in nonwesternized Melanesians. Lipids. 1996 Feb;31(2):153-8.

[5] Lindeberg S, Vessby B. Fatty acid composition of cholesterol esters and serum tocopherols in Melanesians apparently free from cardiovascular disease – the Kitava study. Nutr Metab Cardiovasc Dis. 1995; 5: 45-53.

[6] Xue C et al. Consumption of medium- and long-chain triacylglycerols decreases body fat and blood triglyceride in Chinese hypertriglyceridemic subjects. Eur J Clin Nutr. 2009 Jul;63(7):879-86.

[7] Assunção ML et al. Effects of dietary coconut oil on the biochemical and anthropometric profiles of women presenting abdominal obesity.  Lipids. 2009 Jul;44(7):593-601.

[8] Gao Z et al. Butyrate improves insulin sensitivity and increases energy expenditure in mice. Diabetes. 2009 Jul;58(7):1509-17.

Protein, Satiety, and Body Composition

A number of studies have found protein to be the most satiating macronutrient, with fat moderately satiating, and carbs least satiating.

Thus, when people reduce carbs and increase protein, their appetite declines and they almost always reduce calorie intake. This can leads to rapid short-term weight loss. This is why most popular weight loss diets are high in protein: increasing protein causes dieters to quickly lose some weight, encouraging them to continue.

A 2005 editorial in the American Journal of Clinical Nutrition summarized the evidence that higher protein intake is helpful for weight loss:

The higher than usually recommended protein content of many popular diets, such as the Atkins Diet, The Zone, and The South Beach Diet, seems to point at possible solutions to the obesity epidemic. Many national dietary guidelines have, until recently, recommended that only 10–20% of the calorie content of the diet come from protein; however, 30–40% of the calorie content in the aforementioned diets comes from protein, at the expense of carbohydrates. Newer research indicates that the high-protein content of these diets may actually be the reason for their partial success in inducing weight loss, despite no restrictions in total calories (2)….

In this issue of the Journal, Weigle et al (3) showed that an increase in dietary protein from 15% to 30% of energy and a reduction in fat from 35% to 20%, at a constant carbohydrate intake, produces a sustained decrease in ad libitum calorie intake and results in significant weight loss….

Weigle et al’s results clearly showed that protein is more satiating than is fat, and previous studies have shown that protein is more satiating than is carbohydrate (4). Moreover, diets with a fat content fixed at 30% of calories produce more weight loss when high in protein (25% of energy) than when normal in protein (12% of energy): 9.4 compared with 5.9 kg after 6 mo; after 1 y, evidence was found to suggest that the high-protein diet, independent of the loss of total body fat, resulted in a significant loss of visceral fat (5). [1]

But there are downsides to high protein consumption. Various animal experiments have found that longevity is increased with protein restriction. Also, protein restriction promotes autophagy, which enhances immunity to intracellular bacteria and viruses. So higher protein intake may shorten lifespan and increase the risk of disease.

In a post on his blog (linked in this comment), Dennis Mangan introduced us to the “protein leverage hypothesis.” This hypothesis is put forward in a 2005 paper by SJ Simpson of Oxford University and D Raubenheimer of the University of Auckland [2].

The Satiating Power of Protein

The paper has some graphs which neatly illustrate the satiating power of protein. When animals are given a food formula with a lower protein fraction, they eat more total calories.

Here are some data from rats (b) and chickens (c). The numbers are in kiloJoules; divide by 4.18 to get calories. The animals were on feed formulas with a constant fat content, but different carb-protein ratios. Each data point represents a diet with a different P:C ratio.

If both macronutrients were equally satiating, then the animals would eat the same amount of calories regardless of their food’s protein-carb ratio. The data points would fall on a 45º line (say, for chickens, a constant 1000-kJ line connecting the 1000 kJ mark on the y-axis with the 1000 kJ mark on the x-axis).

But they don’t:  if a line were fitted to these points, it would be much closer to vertical than 45º. The rats, for instance, eat around 150 kJ protein and 75 kJ carb if given high-protein food, but 75kJ protein and 300 KJ carb if given high-carb food. That’s 225 kJ (54 calories) on a high-protein diet, but 375 kJ (90 calories) on a high-carb diet.

The chickens and rats act like protein dominates appetite control:

  • A shortage of protein makes them hungry, and it takes a lot of carbohydrate to satisfy that hunger. So they eat a calorie excess.
  • An excess of protein satisfies their hunger and causes them to quit eating while they are still in calorie deficit.

Evidence in Humans

The same sort of thing happens in humans:

Results a, b, and c are from “short-term” experiments that varied from 2 days to 6 months in length. Results d, e, f, and g are from “long-term” experiments.

People tend to gravitate toward a protein intake of 1520 kJ (360 calories). This can be construed as the “normal” human protein intake, and tends to occur near a carb+fat intake of 8000 kJ (1900 calories). So the “normal” protein fraction of the diet is 360/2260 or 16%. This is consistent with epidemiological data, which finds that nearly everyone worldwide eats near 15% protein.

A line fit to the data has the same steep slope as the animal experiments, but note something interesting. The short-term experiments have a very steep slope, but the long-term experiments have a slope much closer to 45º.

This has to happen. Otherwise, a high-protein diet would lead to permanent calorie deficit which, over time, would lead to starvation. A low-protein diet would lead to permanent calorie excess which, over time, would lead to obesity.

Since we know people neither starve nor become obese due to small adjustments in protein fraction, they must adjust their calorie intake. In the long run, protein no longer controls calorie intake. So there is great protein leverage in the short-term, but much less protein leverage in the long term.

Simpson and Raubenheimer try to develop protein leverage into a theory of obesity. It’s not a very good theory, so I’ve relegated it to an appendix.

Instead, I’d like to talk about what this satiating power of protein means for Perfect Health Dieters.

Implications for Perfect Health Dieters

We have a fairly broad healthy protein range, 200 to 600 calories per day, which brackets the “normal” protein intake of 360 calories. What happens if you shift from 360 calories protein to either the low-protein or high-protein ends of the range?


At low-protein intake, your appetite goes up and total calories go up. You gain a little weight, in the form of adipose mass. This causes leptin levels to increase. As we discussed in “How Does a Cell Avoid Obesity?”, higher leptin (a) lowers appetite and (b) increases thermogenesis, or destruction of fat as waste heat.

Adipose mass increases until the actions of leptin counterbalance the influence of protein leverage.

You reach equilibrium at a slightly higher fat mass and slightly higher leptin levels than on the “normal” protein intake.


At high protein intake, appetite goes down and total calories decrease. You start to lose adipose mass. This causes leptin levels to go down. This (a) increases appetite and (b) decreases thermogenesis, or heat generation.

Adipose mass decreases until a new equilibrium is reached. Equilibrium is reached at a slightly lower weight and slightly lower leptin than on the “normal” protein intake.


The main effect of changing the protein content of the diet is a modest change in body composition.

  • High-protein diets make you leaner and a little lighter.
  • Low-protein diets give you a slightly higher adipose reserve and make you slightly heavier.

The effect is probably small; probably just a few pounds either way. But if you’re looking for to win a bodybuilding competition and you have to become extremely lean and “cut,” you’d do well to adopt a high-protein diet.

It’s probably not a surprise, then, that people with the leanest bodies tend to be healthy but high-protein dieters. Here’s a picture of Anthony Colpo:

I think Anthony has a healthy body, but I don’t think you need to be this lean to be healthy. He would be equally healthy with a few more pounds of adipose tissue.


In the book we say that higher protein intake makes it easier to add muscle, and thus that it may be favored by athletes. Based on today’s post, we can adduce two other reasons to eat a high protein diet:

  1. A more chiseled body. If you want a lean, “cut” look, like Anthony Colpo, high protein will help.
  2. A controlled appetite. In a recent post, Don Matesz stated that he liked a high-protein diet because it helped him auto-regulate his calorie intake. If your goal is “effortless” (willpower-less) calorie restriction, then high protein may help – at least for a while.

However, there are reasons to restrict protein as well. Lower protein intake is likely to extend lifespan, and can increase immunity against intracellular bacteria and viruses, which are behind many late-life diseases.

Is it possible to achieve a lean, muscular body while still gaining the longevity and immunity advantages of low protein intake? And can one lose weight comfortably without assistance from a high-protein diet?  Those will be the topics of Thursday’s post.


[1] Astrup A. The satiating power of protein—a key to obesity prevention? Am J Clin Nutr. 2005 Jul;82(1):1-2.

[2] Simpson SJ, Raubenheimer D. Obesity: the protein leverage hypothesis. Obes Rev. 2005 May;6(2):133-42.

Appendix: The Protein Leverage Hypothesis as a Theory of Obesity

To the satiating power of protein, the protein leverage hypothesis adds two premises:

  1. That any increase in total calorie consumption leads to weight gain which induces insulin resistance in the liver, which in turn upregulates gluconeogenesis. Contrariwise, any decrease in calorie consumption reverses insulin resistance in the liver and downregulates gluconeogenesis.
  2. That the loss of protein associated with gluconeogenesis is treated by the brain’s appetite control centers exactly the same as a decreased intake of protein, and therefore that ongoing gluconeogenesis increases appetite immensely.

The theory of obesity is that once someone starts eating a low-protein diet, their appetite goes up. So they eat a larger amount of total calories, and gain weight. The weight gain causes them to become insulin resistant in the liver. Once that occurs gluconeogenesis is no longer inhibited by insulin, and the liver converts protein to glucose willy-nilly. The loss of protein stimulates appetite. But the person has to eat a lot of excess calories to get enough protein to replace the protein lost in gluconeogenesis. So weight goes up even more. There is a vicious spiral.

If these premises were correct, then:

  • Weight would be unstable. Weight would spiral out of control upward if people ate low-protein diets, and people would wither away once they started eating high-protein diets.
  • Low-carb diets would be extremely obesogenic. Every 1 calorie reduction in carb intake below the body’s daily needs of 600 calories would induce the eating of an extra 1 calorie of protein for purposes of gluconeoegenesis, and on the order of 4 extra calories of fat (by the leverage hypothesis: the P:F ratio stays constant). So each reduction of carb intake by 1 calorie leads to an extra ~5 P+F calories and an increase in total energy intake of 4 calories. Zero-carb diets would induce ravenous appetite, consumption of an extra 3,000 calories per day above the amount needed for weight stability, and obesity and metabolic syndrome would rapidly follow.

Neither is the case.

Mice Who Tear Their Fur Out and The Psychiatrists Who Treat Them

Chris Highcock of Conditioning Research mentioned a fascinating paper yesterday, and then Dr. Emily Deans blogged about it. The paper tells about mice who tore their fur out – akin to the condition of “trichotillomania” in which humans tear their hair out – after being put on a high-tryptophan diet. [1]

Dr.Deans points to the paper’s importance:

As far as I know, it may be the only paper showing a definitive development of psychopathology with an adjustment of diet.  So that’s a big deal!

Since I suspect that most psychopathologies are induced by diet in the context of infection, I think this shows that psychiatric researchers have barely begun to understand their diseases.

As soon as I saw Chris’s post I knew I had to blog about it, because I had similar symptoms to these mice.

My Experience

Briefly, I had a chronic bacterial infection of the brain and nerves, probably from Chlamydophila pneumoniae, plus a few other problems which masked the bacterial infection until I fixed my diet.

C. pneumoniae is a parasitic intracellular bacteria whose main activities are reproduction and diversion of the immune system. Its main effects are:

  • Neuronal hypoglycemia. C. pneumoniae steals glucose products like pyruvate for energy. This can create hypoglycemia in neurons even if blood glucose levels are normal.
  • Serotonin deficiency. C. pneumoniae steals key amino acids like tryptophan, tyrosine, and phenylalanine for protein and niacin synthesis. Of these tryptophan is most important. To block C. pneumoniae activity, the innate immune response triggered by interferon gamma sequesters tryptophan. This denudes neurons of the neurotransmitter serotonin, which is made from tryptophan.
  • Inflammation. C. pneumoniae is able to trigger inflammation which re-directs the immune response away from itself toward extracellular pathogens.

Thus common symptoms of a bacterial infection of the brain are those of cognitive hypoglycemia and serotonin deficiency. Symptoms include:

  • Hypoglycemia : Feeling nervous or jittery; mood changes such as irritability, anxiety, restlessness; confusion, difficulty in thinking, and inability to concentrate; poor coordination.
  • Serotonin deficiency: Anxiety, depression, impaired memory or cognition, low self-esteem, loss of pleasure, poor impulse control, insomnia.

These lists don’t fully capture the experience however.

I started having these symptoms in 1992 during a year-long course of antibiotics, and they would get worse for about the next 15 years. I experienced a dramatic loss of happiness and positive emotions. I had always been happy; now suddenly I wasn’t. Along with this came a weird mental state which is hard to describe, because it has no normal analog. Irritability or anger come closest, so I’ll use those words. But understand that it was a generalized state, not irritation or anger directed at anyone in particular; being naturally phlegmatic, I doubt in 20 years I was uncivil to anyone on more than a few occasions. It was just a persistent irritated/angry emotional state that I was well aware was unnatural and could consciously control.

It seemed like this negative emotional state would build up, and could be discharged a bit by a few expressive habits. I would wring my hands; I still have some slightly twisted finger bones and calluses from over a decade of hand-wringing. And, when alone, I would sometimes scratch my head. This sometimes led to hair loss and bare patches.


This kind of behavior turns out to be not that rare. About 4% of the population is said to have “trichotillomania,” compulsive pulling or twisting of the hair causing hair loss. Trichotillomania strikes women more frequently than men. [Wikipedia, “Trichotillomania” ]

Serotonin depletion is a common feature of mood disorders. I wouldn’t be surprised if most of these disorders are due to brain infections, and the serotonin deficiency is due either to theft of tryptophan by bacteria or to the immune response to intracellular infections, which increases interferon gamma and decreases serotonin.

Evidence, such as it is, is consistent with that idea. People with mood disorders or depression are far more likely than normal people to test positive for antibodies to chronic intracellular pathogens like coronaviruses. [2]

Drugs Help At First, But Often Do Long-Term Harm

The first impulse of modern medicine is to fight the body’s response to disease. If the body has downregulated serotonin, doctors look for drugs that upregulate it.

That is why people with depression and mood disorders are commonly given SSRI’s, drugs that raise serotonin levels.

If these diseases are due to infections, then we would expect the SSRI’s to improve mood immediately, but also to defeat the body’s immune response, supply the pathogens with tryptophan, and promote their replication. As a result, the disease should progress faster. In time, the patient will become worse than would have been the case without the drugs.

And, more often than not, this is what actually happens. Drugs are often “unsafe at any dose”. Antidepressant treatment increases mortality in men by 30%.

The Mice Who Tear Their Hair Out

One of the common breeds of mice used in laboratory research is the C57BL/6 breed. This breed has “an easily irritable temperament … [and] a tendency to bite … [and] display barbering behavior.” [Wikipedia, “C57BL/6”] In barbering, “individuals pluck whiskers and/or fur from their cage-mates and/or themselves.” [1]

C57BL/6 mice also have a modified immune response:

The immune response of mice from the C57BL/6 strain distinguish it from other inbred strains like BALB/c. For example the immunological response to the same pathogen in C57BL/6 mice is often of an opposite spectrum compared to BALBb/c mice, namely C57BL/6 shows Th1 and BALB/c shows Th2 response in response to intracellular pathogen Leishmania major, where a Th1 response results in a resistant ie healer phenotype (since the pathogen is intracellular), whereas a Th2 response results in a susceptible (nonhealer) phenotype. [Wikipedia, “C57BL/6”]

This Th1 response increases interferon gamma levels:

The Th1 response is characterized by the production of Interferon-gamma … [Wikipedia, “Adaptive Immune System”]

Interferon gamma, of course, sequesters tryptophan and diminishes neuronal serotonin levels.

All this sounds familiar: C57BL/6 have lower serotonin; they become irritable and will bite and tear fur out.

Like trichotillomania in humans, tearing of fur is more common in female mice than males:  “Barbering is more frequently seen in female mice; male mice are more likely to display dominance through fighting.” [Wikipedia, “C57BL/6”]

Research Idea: Treat the Mice As We Do Humans

If these mice went to a human psychiatrist (and had health insurance), they’d be prescribed SSRIs to raise their serotonin levels.

A group at Purdue led by professor of animal sciences Joseph Garner decided to see if they could cure barbering through an alternative dietary therapy that would raise serotonin levels just like SSRIs.

[W]e wished to test the hypothesis that a diet which increases serotonin metabolism would decrease the hair-plucking behavior of barbering mice. [1]

The treatment diet was essentially identical to the control diet, except for these differences: Tryptophan levels were four times higher, methionine levels were the same, and other amino acids were halved. Overall protein levels were cut from 24% to 13.3% of calories. Since tryptophan competes with other amino acids for entry to the brain, this shift in amino acid composition led to much larger tryptophan entry to the brain. [1, Table 2] The lost protein calories were made up by increasing carb intake from 57.3% to 68.0%, which I consider a relatively marginal change. In both diets fructose was minimal, 2.5% of calories, and glucose, mostly from starch or dextrose, provided the bulk of the carbs.

The tryptophan was converted to serotonin in the brain, but not for long. Serotonin levels were 55.5 ng/ml in brains of mice on the control diet, 57.6 ng/ml in brains of mice on the high-tryptophan diet [1, Table 3]. However, levels of serotonin metabolites – the leftovers after serotonin destruction – were much higher in the treatment mice.

The results weren’t good:

[E]levating brain serotonin metabolism by tryptophan and carbohydrate supplementation increased the severity of barbering, and induced ulcerative dermatitis. In humans, the induction of compulsive skin-picking by serotonergic agents (SSRIs) has been reported. (24,43) Thus, the current data suggest a homologous outcome in mice, achieved nutritionally instead of pharmacologically. [1]

If you don’t like scientific-ese, here’s Professor Garner in the press release:

[The] diet … was expected to reduce abnormal hair-pulling. Instead, mice that were already ill worsened their hair-pulling behaviors or started a new self-injurious scratching behavior, and the seemingly healthy mice developed the same abnormal behaviors….

“We put them on this diet, and it made them much, much worse,” Garner said.

This does indeed sound like a “homologous outcome” to the experience of human patients treated with SSRIs!

Very likely if the mice had been interviewed at Day 1 after initiation of the high-tryptophan, they would have reported mood improvements, just as human patients do on SSRIs. As with humans on SSRIs, the negative effects took some time to appear. The increase in scratching behavior was not apparent at 6 weeks after initiation of the high-tryptophan diet, but was apparent at 12 weeks (3 months) [1, Figure 4]. Ulcerative dermatitis tended to appear after about 10 weeks on the high-tryptophan diet [1, Figure 3]. 


I’ll follow up in my next post, on Monday, with speculation about what is happening in these mice.

In the meantime, I think it is worth remarking how an intervention thought to be beneficial – restoring serotonin levels to “normal” – has health-impairing consequences over time in both mice and people.

In many ways, contemporary medical practitioners resemble the Sorcerer’s Apprentice. They have at their disposal powerful magic drugs, whose long-term consequences they do not fully understand. The drugs come into wide use, and only years later do we learn that they do more harm than good. And the data showing they don’t work is always “surprising” and “paradoxical.”

In my view, the philosophy behind drug-based medicine is misplaced. Too often drugs are designed to fight or defeat the body’s natural mechanisms. As my parable argued, I believe it is much more effective to cooperate with the body through diet and nutrition.


[1] Dufour BD et al. Nutritional up-regulation of serotonin paradoxically induces compulsive behavior. Nutr Neurosci. 2010 Dec;13(6):256-64.

[2] Okusaga O et al. Association of seropositivity for influenza and coronaviruses with history of mood disorders and suicide attempts. J Affect Disord. 2010 Oct 26. [Epub ahead of print]. Hat tip Dr. Deans,