Addendum to “The Curious Case of Campbell’s Rats”

Several readers forwarded me a response to my post “The Curious Case of Campbell’s Rats” that had been posted on the vegetarian site, 30 Bananas a Day!, and suggested I make a rebuttal.  The response can be found here, and I will quote the relevant portion in full:

Re: Masterjohn’s (MJ) article, since I do not have access to many of the articles he cites, at the moment I cannot assess how accurately he describes all of them. Based on the studies I was able to access and what I could gather the abstracts of others, it seems for the most part the rats having problems MJ attributes to ‘protein deficiency’ (stunted growth, increased susceptibility to aflatoxin poisoning etc.) were weanlings. MJ neglects to mention that although adult rats and humans have similar protein requirements calorically (~5%), rats have considerably higher protein requirements when nursing and as weanlings (~20%), the latter being a period of rapid bodily growth. Indeed rat mother’s milk contains 20% calories from protein vs. 6% for human milk. So no wonder the weanlings were having problems on a 5% protein diet, they really were protein deficient! This would also explain the case of weanling rats successfully blocking aflatoxin initiation fed 20% protein, they were being fed an adequate amount of protein for their age, and hence were better at fighting off disease than their protein-starved counterparts. In any case this oversight of MJ’s is quite misleading.

I certainly agree that protein-deficient diets will only cause stunted growth when they are fed to growing rats.  The younger the rats are, moreover, the more rapidly they are growing, so the effects on growth will be worst when the deficient diets are fed to weanling rats.

The question here, however, is not whether rats eventually grow to an age where they can tolerate low-protein diets.  The question is whether the same low-protein diets that prevent the growth of pre-cancerous lesions once they are formed also increase the toxicity of aflatoxin and other harmful chemicals, decrease the capacity for tissue repair, increase the risk of dying from chemical overdose, and promote the formation of pre-cancerous lesions when these diets are fed to the rats of the same strain, sex, age, and protein requirements.  As we will see below, this is exactly the case.

Estimating protein requirements is extremely difficult.  While it is simple to determine the amount of protein that maximizes the growth of young rats, maximal growth does not necessarily mean maximal health, as Campbell frequently points out.

The most common method of estimating protein requirements in adults is the nitrogen balance method, which is when researchers find the amount of protein necessary to make sure the rats (or people, or other animals) are not excreting more nitrogen than they are consuming in their diets.  This is actually a poor way of estimating protein requirements because it completely ignores how the protein intake affects metabolic functions, like the synthesis and degradation of essential proteins.  The turnover of proteins and the excretion of nitrogen rapidly declines on a low-protein diet, so the optimal intake of protein is almost certainly higher than that required to maintain nitrogen balance.  These facts are acknowledged in nutritional textbooks (this one, for example), and the widespread use of the nitrogen balance method simply reflects the difficulty of reliably estimating protein requirements.

It is true that adult rats require much less protein to maintain nitrogen balance than young, growing rats require to maximize growth.  For example, Imai (2003) found that diets containing 9% casein will maintain nitrogen balance in adult rats while diets containing 25% casein were necessary to maximize nitrogen balance and growth in 4-week-old rats.  Rats are considered “adults,” however, at six months.  Most of Campbell’s rat experiments were conducted using rats much younger than this, so the protein requirement of adult rats is irrelevant.

Appleton and Campbell (1983) showed that 20% casein diets promoted the growth of pre-cancerous lesions when they were fed to rats already dosed with aflatoxin, but protected against the formation of pre-cancerous lesions when they were fed to rats before, during, and soon after the dosing period.  The dose of aflatoxin was equivalent to just under two million peanut butter sandwiches containing 100 grams of peanut butter contaminated with the maximum limit of aflatoxin set by the US government.

Were the rats “adults” during the promotion period?  Not quite.  All the rats were fed 20% casein until they weighed 80 grams.  According to this vendor, male Fischer 344 rats weigh 80 grams when they are just over five weeks old.  The initiation period lasted from five to nine weeks, and the promotion period lasted from nine to 21 weeks.  They would have become adults after another month, but the promotion period corresponds primarily to adolescence.

Were these rats’ protein requirements lower during the promotion period?  Let’s take a look at their growth:

Rats had similar protein requirements during the initiation and promotion periods, even though low-protein diets protected against the promotion and pre-cancerous lesions already formed but promoted the initiation of those lesions.

As we can see above, the low-protein diets affected the final body weight of the rats similarly whether they were provided during the initiation period (red bar) or the promotion period (green bar).

If we consider 5% casein as the reference (the blue bar), providing 20% casein during the initiation period (red bar) increased body weight by 72 grams, while providing 20% casein during the promotion period (green bar) increased body weight by 66 grams.  This difference is so small as to be meaningless and it is not statistically significant.  If we consider 20% casein as the reference (purple bar), providing 5% casein during the initiation period (red bar) decreased body weight by 49 grams, while providing 5% casein during the promotion period (green bar) decreased body weight by 43 grams.  Again, the difference is obviously meaningless and is not statistically significant.

After observing how the red and green bars in the above graph are nearly identical, suggesting similar protein requirements during the initiation and promotion periods, let’s refresh our memories by taking another look at the occurrence of pre-cancerous lesions in the same rats:

Although the rats has similar protein requirements during the initiation and promotion periods, the high-protein diet promoted the growth of pre-cancerous lesions during the promotion period but prevented the initiation of pre-cancerous lesions during the initation period.
The red and green bars don’t seem so close together anymore.  When these two graphs are considered together, it becomes extremely unlikely that the difference between the effects of protein during the initiation and promotion periods is an illusion created by different protein requirements of the rats during the two different periods.  Rather, it appears that high-protein diets promote the growth of pre-cancerous lesions that have already been formed, but provide equally dramatic protection against the formation of these lesions in the first place.

To provide further evidence that the negative effects of the low-casein diets were not limited to very young animals with unusually high needs for protein, Campbell found that the apparent loss of the ability to repair damaged tissue occurred when the 20% casein diet was fed during the promotion period, the same period during which it protected against the growth of pre-cancerous lesions.   Schulsinger, Root, and Campbell (1989) wrote the following:

Under such conditions [wherein low-protein diets suppress the activity of detoxification enzymes], the parent compound AFB1 would accumulate to cause greater acute toxic effects, whereas less product (i.e., aflatoxin-DNA adducts) would accumulate to initiate carcinogenesis. Although this was our previous explanation for the differences between AFB1-induced toxic and carcinogenic responses to protein intake, it would not appear to account for the greater toxic effects observed in the animals fed 20% casein during initiation (when AFB1 metabolism occurs) but 5% casein after initiation. This observation suggests that the low protein intake was not sufficient to allow for tissue recovery from the acute toxic effects.

And thus ended the final paragraph of this paper. Campbell likes to cite it as having showed that plant protein doesn’t promote cancer, but he usually leaves out not only the little part about lysine making plant protein just as active as casein, but also this part about the low-casein diets destroying the capacity for tissue repair.

Both of these studies involved an “initiation” period that preceded and coincided with an acute dose of aflatoxin and a “promotion” period that followed.  This reductionist model is useful for studying the effects of initiation and promotion separately, but it is extremely unrealistic.  None of us will eat two million highly contaminated peanut butter sandwiches over a short period of time, but we are all exposed to small amounts of carcinogens on a day-to-day basis.

In more realistic models where the rats were dosed with smaller (but still large) amounts of aflatoxin every day, the low-protein diets proved fatal, even in adulthood.

Madhavan and Gopalan (1968) fed rats a daily dose of aflatoxin with either 5% (LP) or 20% (HP) casein.  They carried on the experiment for two years, roughly a full lifetime for a lucky lab rat, but they stopped the aflatoxin after six months (when the rats were seven months old, and thus mature adults) because the low-protein rats were keeling over dead left and right:

About 50% of the animals on the LP diet died before the cessation of toxin administration in all experiments and the livers of these showed typical lesions, namely, parenchymal necrosis, bile duct proliferation and fatty change. The majority of the remaining animals died between 52 and 104 weeks with a greater mortality in the LP groups. The death in all was almost invariably associated with a massive pneumonia which was characterized by the presence of microabscesses and mucus-containing cysts.

This is the obscure Indian study with which Campbell had been so enamored.  What of Campbell’s own studies?  Mainigi and Campbell (1980) tried feeding 20% casein and 5% casein (5C) to rats, spiked with 5 parts per million (ppm) aflatoxin as a daily dose, but they had to feed the low-casein group only 2.5 ppm instead:

The 5C animals were fed half the dietary AFB1 since 5 ppm was found to be lethal for this dietary group.

Thus we see that, regardless of the exact protein requirement of adult rats or of humans, the very same low-casein diets that promoted the growth of pre-cancerous lesions once they were formed — by dosing the rats with a couple million highly contaminated peanut butter sandwiches worth of aflatoxin — also increased the toxicity of aflatoxin and protected against the formation of pre-cancerous lesions.

Rather than condemning animal protein or even protein itself, we should be asking whether there is an intermediate level of protein that maximizes our protection against the small doses of toxins to which we are exposed every day, and whether other nutrients found in whole foods rich in protein allow us to reap the protective effects of protein safely.

Read more about the author, Chris Masterjohn, PhD, here.

5 Responses to Addendum to “The Curious Case of Campbell’s Rats”

  1. Wenchypoo says:

    My guess is that they would extrapolate VEGETABLE protein to be that intermediate protein.;)

  2. Hi Wenchypoo,

    Thanks for writing. I did not say anything about an intermediate *type* of protein. I speculated that there might be an intermediate *amount* of protein that maximizes protection against cancer. Animal or vegetable has nothing to do with it.


  3. Stuqua says:

    I suspect that the optimal (but not necessarily intermediate) amount of protein would be whatever most protected against the formation of pre-cancerous lesions to begin with, but that even this amount would be powerless to protect against the artificial induction of such lesions by colossal quantities of aflatoxin.

    Eating two million maximally intoxicated peanut butter sandwiches is something I’ll never do of my own volition. Think of all the gluten in those sandwiches.

  4. D. says:

    It has been my sad experience with vegans (and a few vegetarians, too) that you will not change them, so it’s best to just let them believe whatever they wish. They will keep arguing because they are angry because they are hungry! ;o)

    It’s good to get clarification for those of us who are not vegan or vegetarian, however.

    My other experience with vegans and vegetarians is they somehow think of themselves as superior because they do not “harm animals”. What a ridiculous notion. Even vegetarians are dependent on the animal world, no matter how they want to try to shake it down. These people tend to be haughty; they feel it’s ok to be cruel with words.

  5. JB says:

    to understand that you don’t do a study on powdered food isolates and then attach those results to the whole food. That just reeks of poor science; I knew that when I first started to study nutrition! Why people are so desperate to defend Campbell’s shoddy work is beyond me. I suppose it goes back to the very tired argument of “my dietary theory is better than your dietary theory”.

Leave a reply

© 2015 The Weston A. Price Foundation for Wise Traditions in Food, Farming, and the Healing Arts.