Orientalai*LT Siamai ir Orientalai

LT | EN

Lesson of the Pottenger's Cats
experiment: cats are not humans

In the period 1932-1942, Dr. Francis Pottenger conducted a series of feeding experiments on cats investigating the effects of cooked food. The papers published from that research have been compiled and edited and released as Pottenger [1995]. Pottenger [1946] is relevant as well, and is available as a reprint from the Price-Pottenger Nutrition Foundation.

Dr. Pottenger observed that cats fed a cooked diet developed a number of pathologies, some of which were remarkably similar to certain diseases of civilization, whereas cats eating raw didn't suffer from these problems. It has thus been tempting to blame cooking for all the food-related evils from which we suffer; and in the raw-food movement, this study has been held up as the quintessential paradigm of proof of the perils of cooked food. However, casting Pottenger's experiment in this role suffers from some weaknesses that we shall examine, some of which are fundamental.

    * Domestic cats today reproduce prolifically on "cooked" commercial pet feeds. One obvious remark about the Pottenger's Cats study is that domestic cats all around the world have been living and reproducing successfully for numerous generations on cooked domestic foods, or canned or dry commercial foods. Some cats are so domesticated that they don't even know what to do when they see a mouse, or will play with it for hours (why bother hunting when you are fed with palatable pet food every day?). That is in contrast with Pottenger's cats who couldn't reproduce beyond the third generation. It is certainly true that domestic cats do not have perfect health, but cat nutrition is as complex as human nutrition, and many other factors than cooking determine feline health.

    * Was Pottenger's cooked diet detrimental because it was "dead" or simply deficient? One might argue in response here that commercial feeds are not simply cooked food--they are supplemented as well. But this, then, simply demonstrates it is not that the food is somehow "dead" (as raw-fooders often term cooked food) that is the underlying problem, but rather that the diet fed by Pottenger was deficient in some way. And we shall see later that cooked diets--in humans, at least--are not necessarily more deficient than raw ones. Moreover, one cannot compare pet diets--particularly cat diets--with cooked human diets, which are less monotonous and hence provide a larger variety of nutrients. (Cats are true carnivores, humans are omnivores eating a much wider range of foods.) To put it plainly: cats are cats, humans are humans, and there are significant differences between the two.

      Such considerations, therefore, suggest that--since domestic cats reproduce prolifically on today's cooked-food pet diets (to the point that everyone is urged to spay and neuter their cats to prevent severe overpopulation)--Pottenger's cats suffered not due to some magically bad toxic effects of cooking, or because the food was "dead," but rather from nutritional deficiencies in the diets fed by Pottenger.

    * Study not as well-controlled as newer studies. Pottenger's cat studies were done under limitations that make them less well-controlled than more modern studies. Specifically, many of the cats in the study were donated, and although Pottenger tried to get as much information as possible on the cats' histories [Pottenger 1946], stray cats adopted by people usually come without a reliable health history that cover's the cat's entire life. That is, the likelihood that some of the cats in the study were former strays means the study included cats with an unknown health history, i.e., the use of former strays introduces unknown error in statistical terms.

      Another possible confounding factor is the very nature of the diets used; from Pottenger [1946, p. 467]:

          We placed an order for raw-meat scraps at the market where the Sanatorium meats were bought; these scraps included muscle, bone, and viscera.

      The problem with the above is that it strongly suggests the composition of the diet was not well-controlled, at least not to the level that would be considered essential in modern studies.

      While these two criticisms of Pottenger's methodology do not in themselves invalidate the results of the experiments, they do make inferences drawn from them about cat nutrition less than reliable, since the diet could not have been nutritionally analyzed for key dietary constituents known today but unknown then. This puts significant limits on just what conclusions can be drawn from Pottenger's experiments. Most important in this regard, we shall see in the bullet point just below that the experiments did not anticipate the crucial role of taurine in cat diets, a key nutrient for cats with critical implications for any study of cat nutrition.

    * Pottenger study was done long before the essential role of taurine in cat nutrition was known. Although taurine was discovered in 1838 [DeMarcay 1838, as cited in Huxtable 1992, ref. #134], the modern era of research on taurine did not begin until the late 1960s. See Jacobsen and Smith [1968] for a review of the state of knowledge about taurine at that time, including an interesting discussion of the similarities and differences in taurine synthesis pathways in humans and cats. As well, a comparison of Jacobsen and Smith [1968] with Huxtable [1992] shows the considerable recent expansion of knowledge about taurine, and its importance in mammalian physiology.

      Recall that the Pottenger cat studies were conducted in the period 1932-1942. Although taurine was known at that time, the level of knowledge was much less than today. In particular, the critical role of taurine in cat nutrition--the fact that it is an essential amino acid for cats--was not discovered until 1975, with the publication of two papers: Hayes et al. [1975a,b] as cited in NRC [1986] and Schaeffer et al. [1985]. Inasmuch as the role of taurine in cat nutrition was unknown at the time of Pottenger's research, the possibility that taurine might be a factor in (or explain) the symptoms observed in their cats was obviously unknown to Pottenger and his research team.

    * The cat feeding studies were apparently never replicated or confirmed by any other study. It's worth remarking once again, here, that the reliance on old scientific studies that haven't been updated or confirmed--especially citing a single study without mentioning the context of the rest of the scientific literature on that subject--is a weakness of several other rawist arguments. Sentiment among some raw-fooders seems to be that there must be some secret scientific conspiracy to ignore the Pottenger experiment. But if the experiment has been relegated to the dustbin by science--at least in terms of its applicability to humans--the reason is far more logically the one which follows next.

    * Cats are not an appropriate experimental animal model applicable to humans. Even if one assumes the study to have been a valid one (as far as it went, at least), the decisive and most fundamental criticism of the validity of the Pottenger's Cats experiment is that cats are not used by researchers as an experimental model for humans because the results cannot be extrapolated to human beings with any confidence. More common animal models for which results may have greater relevance for humans are mice, rats, and particularly primates, who are omnivores rather than carnivores as cats are.

    * Differences between cats and humans. Cats, being total carnivores, have special nutritional needs in certain respects when compared to humans (who are omnivores): They require inositol, which is a B-vitamin; but not vitamin C; they need more protein (25-30%) than humans, a fair amount of fat (15-40%), etc. Recall that the SAD is about 15% protein, and that one can live (on a calorie-adequate diet) with 10% protein or even less without visible signs of deficiencies; and that low-fat diets may be as low as 10% fat. It is true that cats and humans do share some metabolic similarities in certain respects as well. (See Metabolic Evidence of Human Adaptation to Increased Carnivory on this site for examples.) However, given the significant differences, drawing more sweeping conclusions about human nutrition from experiments on cats cannot be given much credibility. Most crucially, of particular note are differences in taurine requirements:

          - Cats lack the ability to synthesize taurine and require it in their diet. One of the more striking differences between humans and cats is that the latter require taurine in their diet, which they have no ability to synthesize from precursors. While humans and cats are similar in that both have reduced ability to synthesize taurine in their diets compared to herbivores, the key difference is that cats (who are carnivores) have completely lost the ability and must obtain all that they need of it from their diets. (Humans, as omnivores, have retained the ability to synthesize taurine although it is limited and inefficient compared to herbivores.)

          - Heat-processing negatively affects taurine levels in cats. It has been shown [Hickmann et al. 1990, 1992, Kim et al. 1996a, 1996b] that cats eating heat-processed foods have a lower plasma taurine concentration. The explanation is probably that Maillard reaction products promote an enteric (intestinal) flora that degrades taurine and decreases recycling of taurine by the enterohepatic route. Excessive secretion of the hormone CCK due to a lower protein digestibility might also be another reason [Backus et al. 1995]. (Note: "Enterohepatic" recycling occurs when food absorbed in the lower bowel is transported to the liver for storage and/or processing, some of which is released via the bile back into the upper bowel, where it can recycle again.)

        - Taurine deficiency induces a number of pathologies in cats, such as retinal degeneration, heart disease, reproductive failure, platelet abnormalities, and developmental abnormalities [Waltham 1993, 1994]. In addition, taurine-deficient female cats have taurine-deficient milk [Sturman 1991], which may explain the worsening of the symptoms observed by Pottenger in the offspring.

    * Modern research on inter-generational effects of taurine deficiency in cats. The research of Sturman et al. [1986] is of particular interest in relation to Pottenger's cat experiments because it analyzed the effects of taurine deficiency on pregnant cats and their offspring. In this research, two groups of cats were fed a synthetic diet (obviously also thereby a "cooked" one), except that the diet of one group was supplemented with 0.05% taurine. The female cats were put on the diet for at least 6 months before breeding.

      A wide range of symptoms similar to those Pottenger observed in his cats were observed in the group of taurine-deficient cats in the study by Sturman et al. [1986], as follows.

       - Defects in visual pathway (retinal degeneration) of taurine-deficient cats. Pottenger [1995, p. 10] reported "nearsightedness and farsightedness" in the cats fed cooked food. (Presumably Pottenger did not examine the cat retinas as Sturman et al. did.)

       - Higher incidence of stillbirth in taurine-deficient cats vs. the control group. Pottenger [1995, p. 40] reports higher incidence of stillbirth in his cooked-food cats vs. raw.

        - Much lower survival rate of kittens born to taurine-deficient cats vs. controls. Pottenger [1995, pp. 12, 40] reports lower survival rates in kittens born to cooked-food cats.

        - Kittens born to taurine-deficient cats weighed less than controls [Sturman et al. 1986, p. 658]; compare to Pottenger [1995, p. 40] which reports that kittens from cats on cooked food weighed less than raw.

       - Kittens born to taurine-deficient cats exhibit "neurological abnormalities, including abnormal hind limb development..." [Sturman et al. 1986, p. 656]. Similarly, Pottenger [1995, pp. 10, 12, 27-32] notes nervous system problems in cats receiving cooked foods, differences in long-bone length, and differences in calcium and phosphorus content of cat femurs between the cooked vs. raw groups.

      Another paper, Novotny et al. [1994] notes that taurine deficiency in cats can cause heart disease. Pottenger [1995] does not remark on the cat's hearts, but does report abnormal tissues in the cooked food cat's lungs.

      Note that Pottenger's cats (in the main experiments of interest here) were fed 1/3 raw milk and cod liver oil, plus 2/3 raw meat scraps (raw group), or 2/3 cooked meat scraps (cooked group). Cod liver oil has little or no taurine in it, and cow's milk is extremely low in taurine. Thus the only source for Pottenger's cats to obtain adequate amounts of taurine would have been from the meat scraps in their diet (cooked or raw).

    * Taurine deficiency is a plausible explanation for the symptoms observed by Pottenger. Now consider all the information above, i.e., the considerable similarities between taurine deficiency and the symptoms Pottenger reported for the cats fed his cooked-food diet; the fact that the only source of taurine in the cat's diet was meat; and that heat (cooking) has a negative impact on the bioavailability of taurine to cats. The obvious hypothesis suggested by the above is that the cats in the cooked-food group in Pottenger's research suffered from taurine deficiency. That is, the symptoms of the cooked-food cats were most likely the result of a nutrient deficiency, and not due to "toxins" created by cooking, as suggested by some raw food diet advocates.

      Of course, more than taurine deficiency may be involved here as well; perhaps other deficiencies exist, the only way to know being to exactly repeat Pottenger's experiment, but using various supplementations to understand the true reasons. But modern experiments today, as cited above, make this all basically a moot point. Certainly at the least, they show cats can get along and reproduce well enough on synthetic cooked-food diets--thus demonstrating definitively that cooking itself is not the issue here; something else is.

    * Taurine requirements: once again, cats are not humans. Further, information on taurine in other studies shows that what is valid for cats is not necessarily valid for humans. One interesting thing we know about taurine in humans, for instance, is that vegans who eat no animal products at all have low levels of taurine compared to non-vegetarian humans, whereas conventional (i.e., SAD diet) omnivores have normal levels [Laidlow 1988]. There is an obvious implication of this regarding the effect of cooking. If one makes the assumption that (most all) omnivores displaying normal levels of taurine are ingesting meats or animal foods that are mostly cooked, this suggests that cooking of animal foods does not cause the same problems in humans (an omnivore) as in cats (who are true carnivores). Certainly it shows a typical (mostly) cooked SAD diet does not have the same effect on humans as Pottenger's cooked diet did on cats.

      For more information on taurine, its role in human nutrition, and differences in taurine levels in humans (i.e., vegans vs. the standard Western diet), see the discussion Taurine, a Conditionally Essential Amino Acid (about one-quarter the way down the linked page). Finally, some readers may be curious about the estimated taurine requirements for our feline friends (i.e., the domestic cat). NRC [1986] recommends an intake of 400 mg taurine/kg of body weight for kittens and adult cats, and 500 mg/kg for pregnant female cats.

    * Raw fish blocks thiamine absorption in cats. As an aside, it is known that raw fish contains an enzyme, thiaminase, which blocks thiamine absorption in cats. Thiamine deficiencies may occur if cats are fed raw fish in excess, so we find here a benefit in destroying enzymes by cooking if cats are fed this food. (This point is an issue primarily of academic interest, of course, since fish were an insignificant part of the natural diet of the wild African ancestors of today's domestic housecats.)

    * A further point is that Pottenger's study tells us little about cats fed a more varied diet (Pottenger's cats were fed meat, supplemented with a little milk and cod liver oil), or fed cooked food properly supplemented (in particular with taurine). Importantly, even more, it doesn't tell us anything about 100% raw versus mostly-raw diets, which is one of the primary issues of interest to us here that will be examined later, in Part 3.

Pottenger's cat study was well-conducted for its day, but does not support the usual rawist conclusions

Although a few of the details of the Pottenger cat study might not meet current research standards, it appears that at the time the work was done, Pottenger's study was probably a good one, perhaps even excellent. Despite the major handicap of lack of knowledge about the role of taurine in cat nutrition, Pottenger had the considerable foresight to hypothesize (pp. 19-20, from the reprint of Pottenger [1946]; italics below as in the reprint):

    What vital elements were destroyed in the heat processing of the foods fed the cats? The precise factors are not known. Ordinary cooking precipitates proteins,(7,8) rendering them less easily digested.(9)...

    It is our impression that the denaturing of proteins by heat is one factor responsible.

In summary. Thus we see that the suggestion in this paper--i.e., that the symptoms observed in Pottenger's studies likely were the symptoms of taurine deficiency--was expressed in a less precise form by Pottenger (the only form possible at the time of his research) as a possible explanation. Inasmuch as the above hypothesis of Pottenger appears to be substantially supported by research done 40-50 years later, that is a tribute to Pottenger and his skills as a scientist.

In this connection, we once again note that the rawist idea that "dead" food was responsible for the study results is not only vague but incorrect. Instead, a specific deficiency of a specific nutrient (whether of taurine or possibly some other nutrient(s)) is the explanation required to account for the cats' problems. It should also be noted, once again, that as cats are not a valid experimental model for humans in critical respects such as taurine metabolism, it is invalid to extrapolate the results as having specific relevance for humans. Thus, while Pottenger's results may have been valid (for cats), the usual rawist conclusions about them (for humans) are not.