Question: If it's true that starches such as grains and legumes are a relatively recent addition to the human diet, why then do humans have such an extraordinary ability to secrete the starch-digesting enzyme alpha-amylase, which is present in both saliva and pancreatic secretions? Some biochemists have characterized the level of secretion as "alpha-amylase overkill."

Commentary: The highest levels of alpha-amylase occur in the human pancreas followed by the parotid (salivary) glands. The amylase isozyme levels in parotid glands are of an order of magnitude less than those in pancreas [Sobiech 1983]. Because starch boluses do not remain in the mouth for more than a few seconds, parotid-derived alpha-amylase has little influence upon immediate starch digestion. Additionally, if the starch is wheat-based, there are endogenous alpha-amylase inhibitors in wheat (also in legumes) which effectively inhibit salivary amylase [O'Donnell 1976]. Further, wheat alpha-amylase inhibitors also influence pancreatic amylase secretion [Buonocore 1977] and have been shown to result in pancreatic hypertrophy in animal models [Macri 1977]. Legume starch contains trypsin inhibitors which inactivate native pancreatic trypsin so as to abnormally increase pancreatic cholecystokinin levels and also cause pancreatic enlargement in animal models [Liener 1994].

The point here is that humans obviously have adequate salivary and pancreatic amylase levels to digest moderate amounts of certain kinds of starch. However, antinutrients in our main starch sources (grains and beans) when consumed in excessive quantities may negatively impact endocrine function.

Question: Isn't it true, though, that amylase inhibitors are very heat-labile and denatured by cooking so that they then present little problems in digestion?

Comment: Both alpha-amylase inhibitors (in cereals and legumes) and trypsin inhibitors (primarily in legumes) are not fully denatured by normal cooking processes. It is reported, "Protein alpha-amylase inhibitors may represent as much as 1% of wheat flour and, because of their thermostability, they persist through bread-baking, being found in large amounts in the center of loaves." [Buonocore 1977] Further in his treatise on antinutrients, Liener states, "However, because of the necessity of achieving a balance between the amount of heat necessary to destroy the trypsin inhibitors and that which may result in damage to the nutritional or functional properties of the protein, most commercially available edible-grade soybean products retain 5 to 20% of the trypsin inhibitor activity originally present in the raw soybeans from which they were prepared." [Liener 1994]

Genetic changes in human populations due to agriculture

Question: What exactly are some of the known genetic differences among populations as a result of the spread of agriculture; what is the timescale for any changes; and what is the evidence?

Commentary: It took roughly 5,000 years for agriculture to spread from the Mideast to the far reaches of northern Europe. In this part of the world, agrarian diets were characterized by a cereal staple (wheat or barley early on; later rye and oats), legumes, dairy products, salt, and the flesh of domesticated animals (sheep, goats, cows, and swine). There is strong evidence to suggest that the retention of lactase (the enzyme required to digest lactose in milk) into adulthood is related to the spread of dairying [Simoons 1978]. Most of the world's populations which were not exposed to dairying did not evolve the gene coding for adult lactase retention.

Favism is an acute hemolytic anemia triggered by ingestion of fava beans in genetically susceptible subjects with severe deficiency of glucose-6-phosphate dehydrogenase (G6PD). G6PD deficiency is thought to confer protection against malaria only in those geographic areas where favism exists [Golenser 1983]. A substance in fava beans called isouramil (IU) triggers the hemolytic anemia in G6PD-deficient individuals, and it is this interaction of IU with G6PD erythrocytes which renders these red blood cells incapable of supporting the growth of the malarial pathogen (Plasmodium falciparum). Thus, the spread of agriculture (fava beans in this case) to geographic locations surrounding the Mediterranean was responsible for the selection of G6PD in early farmers.

Celiac disease is an autoimmune disease in which the body's white blood cells (T-lymphocytes) destroy intestinal cells causing malabsorption of many nutrients. The disease is caused by consumption of gliadin (a peptide found in wheat, rye, barley, and possibly oats). Withdrawal of gliadin-containing cereals causes complete remission of the disease symptoms. Only genetically susceptible individuals (certain HLA haplotypes) develop the disease upon consumption of gliadin-containing cereals. There is a geographic gradient of susceptible HLA haplotypes in Europe, with the lowest incidence of susceptible HLA haplotypes in the Mideast and the highest frequency in northern Europe that parallels the spread of agriculture from the Mideast 10,000 years ago. This information is interpreted as showing that agriculture (via wheat, rye, and barley) genetically altered portions of the human immune system [Simoons 1981].

Diseases of insulin resistance, particularly non-insulin-dependent diabetes mellitus (NIDDM) occur in greater frequency in populations that are recently acculturated compared to those with long histories of agriculturally based (high-carbohydrate) diets. It has been hypothesized that insulin resistance in hunter-gatherer populations perhaps is an asset, as it may facilitate consumption of high-animal-based diets [Miller and Colagiuri 1994]; whereas when high-carbohydrate, agrarian-based diets replace traditional hunter-gatherer diets, it (insulin resistance) becomes a liability [Miller and Colagiuri 1994] and promotes NIDDM.

Question: What about D'Adamo's ideas that the ABO blood groups correlate with adaptation to different dietary patterns? Type O is said to be the original hunter-gatherer blood type, with Types A and B having originated more recently in response to agriculture and supposedly being blood types that are more adapted to vegetarian diets.

Commentary: In regard to D'Adamo's ideas concerning ABO blood groups, diet, and disease susceptibility, I suspect that the relationship is significantly more complex than what he has proposed. There are numerous examples in the literature showing an association with blood types and diet-related disease [Hein 1992; Dickey 1994]; however, it is unclear whether a causal relationship is present.

It is generally conceded that human blood types have evolved in response to infectious disease [Berger 1989]. Because there are 30 common blood cell surface antigens (groups) in addition to the ABO group, it seems improbable that if blood typing is associated with certain dietary-induced maladies that they would be exclusively a function of only ABO groups. The two references I have cited demonstrate a relationship with Lewis blood types, not ABO.

Consequently, it is more probable that a complex relationship exists between blood cell surface antigens, diet, and disease that likely involves multiple blood-group types. Further, because of the confounding effect of genetic disequilibrium (the associated inheritance of genotypes that do not follow Hardy-Weinberg equilibrium patterns), the relationship may only be serendipitous in nature and not causal as proposed by D'Adamo.

Evidence of genetic discordance as seen in autoimmune diseases

Up to this point, we have only briefly touched upon the role cereal grains have in inducing autoimmune disease (except for a brief look at celiac disease). There is substantial evidence (both epidemiological and clinical) showing the role cereal grains may play in the etiology of such diverse autoimmune diseases as multiple sclerosis (MS), insulin-dependent diabetes mellitus (IDDM), rheumatoid arthritis, sjogrens syndrome, dermatitis herpetiformis, and IgA nephropathy.

Although this proposal may at first seem preposterous, there is strong data to suggest that cereal grains may be involved in all of these diseases through a process of molecular mimicry whereby certain amino acid sequences within specific polypeptides of the gramineae family are homologous to (have the same structural form as) a variety of amino acid sequences in mammalian tissue. These homologous amino-acid (AA) sequences can ultimately confuse our immune systems so that it becomes difficult to recognize "self" from "non-self." When this happens, T-cells, among other immune-system components, launch an autoimmune attack upon a body tissue with AA sequences similar to that of the dietary antigen.

It seems that grass seeds (gramineae) have evolved these proteins with similarity to mammalian tissue to protect themselves from predation by mammals, vertebrates, and even insects. This evolutionary strategy of molecular mimicry to deter predation or to exploit another organism has apparently been with us for hundreds of millions of years and is a quite common evolutionary strategy for viruses and bacteria. It has only been realized since about the mid-1980s [Oldstone 1987] that viruses and bacteria are quite likely to be involved in autoimmune diseases through the process of molecular mimicry. Our research group has put together a review paper compiling the evidence (and the evidence is extensive) implicating cereal grains in the autoimmune process, and with a little bit of luck it should be published during 1998. [Editorial note as of June 1999: The paper has now been published; the citation is: Cordain L (1999) "Cereal grains: humanity's double-edged sword." World Review of Nutrition and Dietetics, vol. 84, pp. 19-73.]

Without the evolutionary template and without the evidence provided us by the anthropological community showing that cereal grains were not part of the human dietary experience, the idea that cereal grains had anything to do with autoimmune disease would probably have never occurred to us. This new electronic medium has allowed instant cross-fertilization of disciplines which probably would have rarely occurred as recently as five years ago.

How peptides in cereal grains may lead to molecular mimicry and autoimmune disease

In the human immune system, there are a number of individual mechanisms which allow the body the ability to determine self from non-self so that foreign proteins (i.e., bacteria, viruses, etc.) can be recognized, destroyed, and eliminated. Perhaps the most complex system which nature and evolution have engineered to accomplish this is the human leucocyte antigen (HLA) system. This system was discovered when early physicians found out that tissue from one human could not be grafted to another without rejection. The physiological function of this system was not to foil the efforts of transplant surgeons, but to initiate an immune response to parasites (viruses, bacteria).

All cells of the body manufacture HLA proteins, whose function is to bind short peptides (protein fragments) and display them on the cell surface. Most of the peptides are derived from the body's own proteins (self-peptides). However, when the body is infected by a virus or bacteria, the HLA molecules pick up peptides derived from broken-down proteins of the virus or bacteria and present them to T-lymphocytes. The purpose of T-lymphocytes is to continually scan the surfaces of other cells to recognize foreign peptides while ignoring self-peptides.

Once a T-cell receptor "recognizes" a foreign peptide, a complex series of steps is set into play which ultimately destroys the cell presenting the foreign peptide as well as living viruses or bacteria in the body which also have peptide sequences similar to those which were presented. When the HLA system loses the ability to recognize self (self-peptides) from non-self (foreign peptides), T-lymphocytes attack self-tissue, resulting in what is known as an autoimmune disease (i.e., celiac disease, IDDM, multiple sclerosis, dermatitis herpetiformis, ankylosing spondylitis, etc.).

The HLA proteins which present foreign peptides to circulating T-lymphocytes are coded by DNA sequences on chromosome 6. The entire HLA system includes more than 100 genes, and occupies a region more than four million base pairs in length which represents 1/3,000 of the total human genome. On chromosome 6, the HLA is sub-divided into Class I (HLA-A, HLA-B, HLA-C) and Class II segments (HLA-DR, HLA-DQ, HLA-DP). Individuals with autoimmune disease inherit characteristic HLA combinations which identify their disease. People with celiac disease have genetic markers (HLA-DR3, HLA-B8, and HLA-DQ2) which are associated with the disease; people with insulin-dependent diabetes mellitus (IDDM) almost always have DQ and DR genotypes. Thus, the manner in which foreign proteins are presented to circulating T-cells by HLA proteins tends to be different for individuals with autoimmune diseases compared to those without these maladies.

As mentioned previously, the incidence of a variety of autoimmune diseases follows a southeasterly gradient from northern Europe (highest incidence) to the Mideast (lowest incidence). This gradient occurs because the incidence of susceptible HLA haplotypes increases as one moves northwesterly from the Mideast. This gradient--which occurs for both the incidence of autoimmune diseases and HLA haplotypes--is not a serendipitous relationship, but occurred as a result of the spread of agriculture from the Mideast to northern Europe [Simoons 1981]. Consequently, as agriculture spread into Europe there were environmental elements associated with this demic expansion ("demic" means the spread of genes through either the migration or interbreeding of populations) which progressively selected against HLA haplotypes (combinations of HLA genes inherited from the two chromosomes in each cell) that were originally present in the pre-agrarian peoples of Europe.

Now, the question is, what were those environmental selective elements? In the case of celiac disease, it doesn't take a rocket scientist to determine that it was wheat. Increasing consumption of wheat caused increased mortality from celiac disease--thus, the incidence of celiac disease and its susceptible HLA haplotypes (HLA-B8, HLA-DQ, HLA-DR) are lowest in those populations with the most chronologic exposure to wheat (Mideasterners and southern Europeans) and greatest in those populations with the least exposure (northern Europeans). Similar arguments can be made for IDDM and a host of other autoimmune diseases. There are a substantial number of animal studies showing that consumption of wheat by rats increases the incidence of IDDM [Scott 1988a; Scott 1988b; Scott 1991; Elliott 1984; Hoofar 1993; Storlien 1996; Am J Physiol 1980;238:E267-E275; Schechter 1983].

How is it that wheat can wreak such havoc with the autoimmune system? Our research group believes that wheat contains peptide sequences which remain undigested and which can enter into systemic circulation. These peptide sequences are homologous to a wide variety of the body's tissue peptide sequences and hence induce autoimmune disease via the process of molecular mimicry. E.g., macrophages ingest the circulating wheat peptides. HLA molecules within the macrophage then present amino-acid sequences of the fragmented peptide to circulating T-lymphocytes, which through clonal expansion create other T-cells to "attack" the offending dietary antigen and any other self-antigen which has a similar peptide sequence--i.e., the body's own tissues.

The original non-agricultural HLA haplotypes conferred selective advantage in earlier evolutionary times because these genotypes provided enhanced immunity from certain types of infectious diseases. However, with the advent of cereals in the diet they represented a liability. Thus, the genetic data clearly shows that a recently introduced food type has resulted in genetic discordance between our species and those from the gramineae family.

Possible autoimmune connection between dietary peptides and some forms of autism

Autism in children is a neuro-developmental disorder characterized by few or no language and imaginative skills, repetitive rocking and self-injurious behavior, and abnormal responses to sensations, people, events, and objects. The cause of the syndrome is unknown, but there is increasing evidence that it may be autoimmune in nature.

Reed Warren's group [Singh 1993] found that 58% of autistic children maintained antibodies to myelin basic protein (a protein found in the myelin sheaths of nerves and suspected of being the target protein [self-antigen] for T-lymphocytes in the autoimmune disease multiple sclerosis). Additional support for the concept that autism may be autoimmune in nature comes from work showing that 46% of autistic children maintain major histocompatibility complex (MHC) alleles associated with the disease [Warren 1996]. The function of the MHC is to present self- and foreign peptides to circulating T-lymphocytes at the surface of all cells throughout the body. Thus, if foreign peptides are presented by the MHC, circulating T-lymphocytes can mount an immune response on the cell or cells that present, via the MHC, those foreign peptides, and destroy them.

The MHC not only presents foreign peptides, but it also presents peptides derived from the proteins of genes comprising the MHC itself. The susceptiblity genes for autism are: DRB1*0404, DRB1*0401, and DRB1*0101 [Warren 1996]. In a particular portion of these genes (the third hypervariable region [HVR-3]), there is a common amino-acid sequence shared by all three genes. This amino-acid sequence is either QKRAA (glutamine-lysine-arginine-arginine-alanine-alanine) or QRRAA. Thus, either the QKRAA amino-acid motif or the QRRAA amino-acid motif can be presented to circulating T-lymphocytes. This particular shared epitope increases the susceptibility to a number of autoimmune diseases, including rheumatoid arthritis [Auger 1997]. (An epitope is the part of an antigen recognized by an antigen receptor, i.e., a specific amino acid sequence of a protein.)

The QKRAA or QRRAA amino-acid motif also occurs quite frequently in pathogens which reside in the human gastrointestinal tract including Escherichia coli, Proteus mirabilis, Lactobacillus lactis, Brucella ovis, and many other anaerobic gut bacteria [Auger 1997]. The QKRAA or QRRAA sequences are found specifically in a particular type of protein contained in gut bacteria, called DnaJ proteins. DnaJ proteins normally have a bacterial partner/ligand protein called heat-shock proteins (HSP70). It is the QKRAA or QRRAA amino-acid sequence of DnaJ which allows it to bind HSP70.

When the MHC presents endogenously derived DRB1 alleles which contain the QKRAA or QRRAA amino-acid motif, then circulating HSP70 proteins (which normally bind DnaJ proteins) can bind the body's own MHC-presented QKRAA or QRRAA sequences. Circulating CD4+ T-lymphocytes recognize this HSP70/QRRAA sequence as foreign, and mount an immune response on all cells presenting this (HSP70) amino-acid motif.

We believe that myelin basic protein contains an amino-acid sequence that is homologous to an amino-acid sequence found in HSP70, and it is this three-way mimicry between DRB1 peptides, bacterial peptides, and self-peptides which causes self-tolerance to be broken.

So, how does a paleodiet have anything to do with this process? Paleodiets are characterized by their lack of cereal grains, legumes, dairy products, and yeast-containing foods. Both cereal grains and legumes contain glycoproteins (conjugated proteins that have a carbohydrate as the non-protein component) called lectins which bind intestinal epithelial cells and change the permeability characteristics of these intestinal cells [Liener 1986; Pusztai 1993b]. Not only do these lectins cause an increase of the translocation of gut bacteria to the peripheracy, they cause an increased overgrowth of gut bacteria as well as a change in the gut flora [Liener 1986; Pusztai 1993b]. Further, cereal and legume-derived lectins (WGA and PHA respectively) cause increased expression of intracellular adhesion molecules (ICAM) in lymphocytes [Koch 1994] which allows bacterial/immune complexes to move from gut to the affected tissue. Additionally, cereal and legume lectins increase lymphocytic expression of common inflammatory cytokines such as tumor necrosis factor alpha (TNFa), interleukin 1 (IL-1) and IL-6 which are known promoters of autoimmune disease.

The cell walls of cereals and legumes contain a storage protein, GRP 180, which also can act as a ligand to self-presented MHC peptides [Dybwad 1996]. Further, peptides contained in dairy proteins (bovine serum albumins--BSA, among many) also may contain peptide sequences which can interact with endogenously presented peptides [Perez-Maceda 1991]. Cereal, legume, dairy, and yeast-free diets potentially have therapeutic benefit in many autoimmune related disorders via their ability to reduce gut permeability and decrease the exogenous antigenic load both from pathogenic bacteria and from potentially self-mimicking dietary peptides.

Conclusion

The study of the evolutionary relationship of diet to health and the repercussions of evolutionary discordance on human biology is still in its infancy. As an interdisciplinary study that depends on the correlation of insights from numerous scientific fields such as archaeology, anthropology, ethnobotany, evolution, biology, genetics, autoimmunity, and clinical nutritional research, its insights have to this point been appreciated only by a relative few.

However, the case of cereal grains and legumes in the diet presents one of the clearest examples where data from these various disciplines is beginning to come together in a way that can be tested and verified by controlled clinical studies. As such it represents an example of the power of the evolutionary approach to provide new directions for research that can give us insights into human diet that may not have been heretofore possible.

Cereal grains are a particularly good example of both the promise and the perils of the evolutionary process--in both its physiological and cultural guises. Without cereal grains as the agricultural base for modern civilizations, it is exceedingly doubtful whether humanity would have developed the culture or the technologies that have led to the accomplishments and scientific insight we now enjoy. Obviously, modest amounts of cereal grains can be a part of the diets of most people with effects that, at least given our current state of knowledge, can be considered negligible.

At the same time, however, to rely on cereals as more than a supplemental part of the diet increases the likelihood they will lead to problems due to their evolutionary discordance. For some people with the "wrong" genetic heritage (even though it may be a very normal aspect of the inherent variability of the human genome), any cereal grains at all are destructive enough--as in the case of celiac disease--that they are simply not an option if such individuals wish to avoid health problems due to their consumption.

As well, with the advent of the young and newly expanding field of autoimmune research, there is increasing recognition of the role that autoimmunity may play in more disease conditions than has perhaps heretofore been thought. Given the potential that peptides in cereal grains have shown for precipitating autoimmune dysfunction, it may be that linkages with further disease conditions will be discovered as time goes on, which is cause for concern and points to the need for more intensive research in this area.

I hope you have enjoyed our look at this example of the power of the interdisciplinary approach employed in the field of paleolithic diet to bring new insights to the study of human diet and uncover compelling new areas for nutritional research.

--Loren Cordain, Ph.D.

For further Paleodiet research from Dr. Cordain, you can download printable PDFs of his research group's peer-reviewed papers (more than 20 at last count) at his website, plus get information about his 2002 book, The Paleo Diet.

Back to Paleodiet & Paleolithic Nutrition