Wednesday, October 5, 2011

Dietary Protein, IGF-1, and Hyperinsulinemic Diseases of Civilization

In their paper, “Hyperinsulinemic diseases of civilization: more than just Syndrome X,” Loren Cordain, Michael Eades, and Mary Dan Eades argue that dietary carbohydrate intake, particularly of  refined sugars, promotes hyperinsulinemia, which in turn raises levels of insulin-like growth factor-1 (IGF-1) and androgens, while reducing levels of insulin-like growth factor-binding protein-3 (IGFBP-3) and sex hormone-binding globulin (SHBG).[1]

These endocrine system changes create an environment that promotes cell proliferation and growth, acne, early menarche, epithelial cell cancers (breast, prostate, and colon), increased stature, myopia, cutaneous papillomas (skin tags), acanthosis nigricans, polycystic ovary syndrome (PCOS) and male vertex balding.

Effect of Protein Intake on IGF-1: Experimental Results

In “Long-term effects of calorie or protein restriction on serum IGF-1 and IGFBP-3 concentration in humans,” Fontana et al report on their studies of the effect of either caloric restriction or protein restriction on these endocrine markers in humans. [2]

In one study, they compared the effects of three interventions:

1) 20% caloric restriction with no change in energy expenditure (CR)
2) 20% increase in energy expenditure with no change in caloric intake (EX)
3)  healthy lifestyle control group with no specific manipulation of energy balance (HL)

All three groups consumed diets supplying about 16% of energy as protein. 

After one year, although both the CR and the EX interventions produced substantial and similar changes in body mass (~ 6 kg lost) and body fat percentage (~ 23% decrease), neither intervention reduced levels of IGF-1 or increased IGFBP-3.  This contrasts with rodent studies, wherein caloric restriction does reduce IGF-1 and increase IGFBP-3.

In the second study, they compared the IGF-1 and IGFBP-3 levels in members of the Caloric Restriction Society (CR) who had been practicing caloric restriction for an average of 6 years, with those of age-matched controls consuming a typical Western diet (WD).  The CR group averaged only 1800 kcal per day with 24% of energy from protein and 28% from fat, while the WD group averaged 2500 kcal per day with 16% from protein and 34% from fat. 

Thus, the CR group averaged 108 g of protein daily, and the WD group 100 g of protein daily, both at least twice the required amount for a 150 pound male.

They reported: 

“As in our 1-year CR study, we found that there were no differences in serum IGF-1 and IGFBP-3 concentrations, and IGF-1 : IGFBP-3 ratio between the CR and Western diet groups.... These data provide evidence that, in contrast to the decrease in IGF-1 in rodents, a reduction of IGF-1 expression is not a component of the adaptive response to long-term CR in humans.”

In contrast to these findings, Fontana et al note that fasting for 10 days markedly reduces IGF-1 into the range found in growth hormone deficient patients, and this correlates closely with the excretion rate for urea, a marker for nitrogen balance and thus protein intake.

Consequently, Fontana et al did a third study “comparing the serum IGF-1 and IGFBP-3 concentrations, and IGF-1 : IGFBP-3 ratio in 28 vegans who had been consuming a moderately protein-restricted (PR) diet (0.76 g kg−1 per day; ~10% of intake from protein) for ~5 years age-matched with 28 members of the Calorie Restriction Society who consume a high-protein diet (1.73 g kg−1 per day; ~24% of energy intake from protein).” 

The vegan/PR group had an average protein intake of 50 g per day, while the CR group averaged twice as much, 106 g per day.

Fontana et al found that “Both serum IGF-1 concentration and IGF-1 : IGFBP-3 ratio were significantly lower in the moderately PR diet group than in the severe CR diet group, whereas fasting insulin and C-reactive protein were similarly low in the moderately low-protein vegan and CR groups.’

In other words, the vegan/PR group had the most favorable IGF-1 and IGFBP-3 levels, while also having low serum insulin and C-reactive protein comparable to the CR group, without enduring caloric restriction.

This table compares the values of the PR group to the CR group:

The low protein group had an IGF-1 level 21% lower than the low calorie group

Of interest, the “serum total and free IGF-1 concentrations were lower in the moderately PR group than in the severe CR high-protein diet group, despite the PR groups’ higher body weight, BMI and body fat content.”

This raised the possibility that the higher protein intake of the CR group prevented desired declines in IGF-1, so Fontana et al performed a fourth study.  They had 6 CR volunteers to reduce their protein intake from 1.67 g/kg per day to 0.95 g/kg per day, a 43% reduction, while maintaining caloric intake constant.  The 0.95 g/kg/d level of intake still exceeds the reference daily intake of 0.83 g/kg/d, which covers the needs of 98% of the population.

As a result of this reduction of protein intake, the CR volunteers had an average 25% reduction in IGF-1 levels, confirming the hypothesis that high protein intake prevents reduction of IGF-1 levels even in the context of caloric restriction.

This then suggests that high protein intake may raise IGF-1 levels and thus promote all the diseases that Cordain, Eades, and Eade linked to elevated IGF-1 and androgens and reduced IGFBP-3 and SHBG: acne, early menarche, epithelial cell cancers (breast, prostate, and colon), increased stature, myopia, cutaneous papillomas (skin tags), acanthosis nigricans, polycystic ovary syndrome (PCOS) and male vertex balding.

Another question arises from this research:  Since the vegans had lower IGF-1 levels than the omnivorous CR group, and plant proteins have different amino acid profiles than animal proteins, I wonder if plant protein has a different effect on IGF-1 than animal protein?

Who eats a high protein diet?

Most people do not understand that the reference daily intake (RDI) is not a minimum requirement.  In fact, the RDI for protein, 0.83 g/ kg of body weight per day, is calculated to cover the needs of the people in the 98th percentile of requirements.

Many people require substantially less protein than the RDI.  The median requirement for the general population  actually only amounts to  0.65 g/kg per day, about 20% less than the RDI.  [ 4 ]

For a 150 pound (68 kg) male with 10% body fat, this translates to a requirement of 40 to 50 g of protein daily.  NHANES data suggests that half of U.S. males consume 40% more than the RDI level of protein, which means that many consume more than twice as much protein as they actually require.[5

When I ate a meat-based paleo diet, my protein intake ranged from 130 -190 g per day, two to four times the maximum requirement of someone with my bodyweight; which according to human protein requirements research [6] amounts to an excessive protein intake. 

According to this research, I probably had an elevated IGF-1 level and ratio to IGFBP-3. 


cwaiand said...

so,from high fat meat based to sure can flip based on a couple of articles.i,m not saying your wrong but there is no way to know if your right.

we ar right now in the middle of dietary pergatory.noone can tell me what i should eat.i have no clue and i think it,s time to stop reading blogs.

i think i should just copy my wife ,thin healthy eats what ever she wants(white bread ,fruit loops)just controls portions.she has the blood work of an 18 year old and is the same weight as high school.she is 46 years old,she,s never bought into any of the crap of told her(i got it from reading retarded blogs).i,m done.

Don said...


I certainly didn't 'flip' based on just a couple of articles. Comparatively, the vegans have a huge pile of evidence on their side, while the high fat advocates have slim evidence. They wish the Eskimos were healthy...too bad they were the most unhealthy of primitive groups.

There certainly is a way to know if I am on the right track.

Your wife sounds like the perfect example of evidence that humans are designed for a carbohydrate-based diet.

A solid understanding of human nutritional requirements will make sense of people like your wife, as well as the anthropological data and international data and human physiological data.

I left high fat meat-based 'paleo' because I have seen just too much evidence that contradicts the idea that humans are designed for a high fat, meat-based diet, including people like your wife. If the Taubes hypothesis was correct, your wife should be fat and have dyslipedemia.

Anonymous said...


Maximize your Nutrient:Calorie ratio and minimize your Toxin:Calorie ratio.

Now you're not eating to calorie or nutrient excesses and you've eliminated grains, industrial seed oils, and processed sugars.

Ta da. It really couldn't be any more simple. Don's ideas and those of every other "paleo" blogger on the planet haven't been inconsistent with the above either.

I think you have other frustrations/issues as indicated by your perception of supposed "dietary purgatory". Get more sleep, have your cortisol checked, practice meditation, etc etc....

Your wife has remained leptin sensitive. Good for her. She hasn't over ate herself into mitochondrial dysfunction. The same rules *obviously* do not apply for someone who is obese, on their way to obesity, or formerly obese. You should read for more info on mitochondrial dysfunction.

Those people will need to follow one of the five flavors of paleo:

My nod goes to number 5, The Perfect Health Diet. Reading the book you'll find out how to easily design a sustainable and well nourished paleo diet that is free from toxins and excess nutrients while having *reasonable* amounts of protein and fat for your goals...

Anonymous said...


Do you think an obese person with mitochondrial dysfunction could lose weight with a high carb diet?

Isn't the effect of the same food on a lean vs obese person going to be different?

WoLong said...

Excellent post. What kind of evidence did Cordain et al, use to support their claims? I'm not an expert in nutrition science so I can't make sound judgement. I'll try to read their paper as well aa Fontana's.

Another question relates to excercise. Does high intensity exercise lead to elevated IGF-1? If so, should one give up on HIT if the goal is to keep IGF-1 low?

JasonB said...

Does this account for the anecdotal evidence of athletes who note the more protein they eat, the better they perform and recover? It's not Mat Lalonde approved proof, but athletes tend to naturally get a lot of this correct (outliers do exist; don't bring up the Lamar Odoms' of the world as evidence).

Don, what does your physical fitness routine look like?

Don said...


I certainly think that an obese person can lose weight on a diet containing a high proportion of energy from carbohydrate-rich whole foods. I have seen it happen.


I'm not saying Cordain et al are wrong about refined carbohydrates, but that they ignored evidence that protein intake also affects IGF-1.

Cordain and the Eades basically argue this way:

1: Diets rich in refined carbohydrates, particularly those rich in fructose, raise chronic insulin levels may promote hyperinsulinemia.
2: Hyperinsulinemia raises IGF-1 and reduces IGFBP-3 levels.
3: Historical increases in disorders promoted by elevated IGF-1 correlate with increases in refined, high-glycemic carbohydrate consumption.
4: Therefore, high-glycemic carbohydates may be the promoters of these disorders.

They left out the fact that high protein foods also stimulate insulin release and the research by Fontana et al was published 6 years after Cordain and the Eades published their paper.

Again, I'm not saying that refined carbohydrates don't play a role; but I am saying, we actually have experimental human data showing that elevated protein intake increases IGF-1 substantially, even in the context of caloric restriction.

Andreas said...

Hi Don!
Interesting point along the line of our conversation at the AHS. I agree that most evidence point toward excess protein being problematic in the long run. I don't think that dietary fat is harmful by itself, and it might offer some protection against the damage of eating excess protein.

Don said...

Jason B,

You can check my posts on strength training to see my typical routines.

Right now I am not doing any heavy training because I am using the Egoscue method to correct musculoskeletal misalignments that have given me repeated injuries.

Other than strength training, I walk and have recreational activities.

I don't think this accounts for any anecdotal evidence or claims of athletes to perform or recover better eating more protein. Few athletes have done the requisite n=1 experiments to make their claims. For example, they would have to make isocaloric substitutions for protein to ensure no caloric change when eating less protein or animal products.

Since athletes eat more food than non-athletes, so long as they eat whole foods, they would automatically consume more protein than a non-athlete. To use a simple example, 3000 kcal of potatoes provides 50% more protein than 2000 kcal of potatoes.

RawHemp said...


I lost 60 pounds on a diet of mostly WW bread, peanut butter, honey, OJ and quesadillas with the occasional mcdonalds treat or similar. I also made a ton of lifestyle changes.

Martin said...


You mentioned previously that a high fat diet reduces gut motility increasing the risk for constipation. Is there a paper you can link to?

Don said...


Here's one:

To my knowledge, it is widely accepted that fat delays gastric emptying and intestinal transit, and this is one of the reasons people feel so 'full' from high fat meals for long periods after the meal.

Unfortunately for fat, the full sensation that comes after the meal does not prevent a person from consuming an amount of fat exceeding caloric requirements during the meal.

Simply, fats supply 9 kcal/g and starches like potatoes, only about 1 kcal/g. The typical person eats about 1500-2500 g of food daily; this simply reflects the amount of food one must eat to fill up the stomach adequately at a meal to feel full. If composed entirely of potatoes, you have 1500 to 2500 kcal. If composed entirely of nuts, at about 5 kcal per g, you have 7500 to 12500 kcal. In short, the more fat-rich food in the diet, the more likely one will spontaneously eat more food energy than you require.

Peter said...

Don, the high-fat crew not only tries to convince us that Inuits are healthy but that Masai would be that as well.

You can avoid the strokes by walking/jogging 30km per day, but you even that does not rescue you from having arteries of that of old American men.

"Atherosclerosis in the Masai"

dr.addn said...

I'm a big fan of legumes,
so I'm basically on the vegan train already;
but, I like the numbers from the Zone diet:
50g meat, 70g low-glycemic carb,
and then mono'oils to suit your energy preferences;
did you have evidence this was sub-optimal ?

marcus volke said...


: IGF1 levels in the upper normal range are associated with reduced BP (6) and vascular tone (7), increased insulin sensitivity (8, 9), and reduced prevalence of diabetes mellitus(10). Epidemiological studies have suggested that IGF1 levels in the lower normal range are associated with an increased risk of ischemic heart disease (11,12, 13) and stroke (14, 15, 16, 17). In this setting, a protective role in the development of atherosclerosis was suggested for free IGF1 levels (17).

Don said...

Marcus volke,

that study found that people with low normal IGF1 also had high cholesterol and triglycerides:

"The subjects with normal total cholesterol (0.09±0.94 SDS) or triglyceride (0.02±0.97 SDS) levels had significantly higher IGF1 levels than those with hypercholesterolemia (−0.88±1.16 SDS, P<0.0001) or hypertriglyceridemia (−1.13±1.18 SDS, P<0.0001); these latter were similar to each other. IGF1 SDS was significantly correlated with total cholesterol (r=−0.32, P<0.0001) and triglyceride (r=−0.34, P<0.0001) levels. IGF1 SDS did not independently predict total cholesterol (t=−1.41, P=0.17) nor triglyceride (t=−1.80, P=0.073) levels. "

So how does this apply to my case? I had high cholesterol >220 when I ate the high protein animal based diet, and reduced it by reducing animal products. From a quick read of that study, I surmise that the factors that increase risk of stroke, atherosclerosis, HTN, and diabetes also may depress IGF-1 levels. I also find it interesting that they did not provide age group data, but lumped together all subjects aged 18-80. It is well known that GH and IGF-1 levels decline with age, while heart disease, HTN, etc all increase with age. Therefore, it should be no surprise that high IGF-1 (a marker of youth) is associated with a lower risk of age-related diseases. This is basically saying that young people have lower risks of these diseases. Duh.