<<Disclaimer: Verify this information before applying it to your situation.>>

Hi Kemp,

Can we begin by agreeing that the specific mechanism by which gluten increases
the risk of malignancy has not yet been identified?

> I take as given that gluten does cause extra cancer in
>celiacs, but that just like any other chemical or toxic agent, its effect
>scales with the accumulated dose here in gluten gram years.

If that were demonstrably correct, I don't think I would worry about trace
amounts of gluten. I would re-establish a tolerance (as happened in infancy)
and be much less vigilant about the trace amounts. But there is strong
evidence that small amounts may increase our risk as much as a regular diet,
as I outline below.

>Each seed has an equal small
>micro-chance of later leading to cancer: the total probability comes from
>their sum. No repair mechanisms then. The resulting linearity of cancer
>probability with total dose is well established for radiation and benzene,
>for example. Twice the daily dose, twice the chance of cancer; Same dose,
>half as long, half the risk, etc.

And that makes the assumption that dietary gluten is acting directly, as a
carcinogen, in a cumulative manner. I don't think that is a defensible
position.

>Conclusion: if the gluten intake is less than 5% of that in a normal diet
>then  the extra risk of cancer is less than 5% (numerical coincidence) of
>its normal occurrence (citations below). The real risk at such low levels
>and below is instead in nutritional deficiencies. The latter can be
>monitored. Unlike cancer, they are largely reversible: at low doses an
>equilibrium builds up between the regrowth of the microvilli and their
>destruction by whatever gluten is still being eaten.

But there is better evidence that gluten is not, of itself, a carcinogen.
Several notions have been put forward on this topic. One is that chronic,
exponential increases in mitosis, due to gluten-induced damage, result in a
vastly greater risk that mutant cells will occur, and multiply out of
control. If that is the underlying oncogenesis in cd, then your assessment
is likely correct. Identified chromosome instabiliy in untreated celiacs, as
indicated in Fundia, et. al. would appear to support that perspective(1).

But there is other work that contradicts this. Maclauren et al have shown
that natural killer cell activation against Burkitt lymphoma cells is weak
to non-exitant in celiacs. Even celiacs on a gluten-free diet. They showed
no significant difference between treated and untreated celiacs even after
long periods of treatment (2). On the surface, that seems to contradict the
Holmes et. al. study that you cited. A quick glance at the time and place of
the study indicates that it was done in the UK at a time when wheat starch
was allowed.

The gluten content of wheat starch is comparatively small. Perhaps much less
than  the 5% you mentioned. Yet there was no significant difference between
the  reactivity to lymphoma cells, of treated or untreated celiacs. That is
compelling evidence which is contrary to your notion of a linear
relationship between cancer risk and quantity of gluten consumed.

>The reference Holmes et al "Malignancy in coeliac disease--effect of a
>gluten free diet" Gut (1989) 30, 333-338.
>
>The correct comparison/conclusion is quoted even in the abstract: "A
>two-fold relative risk of cancer was found and was because of an increased
>risk of cancer of the..." (various  sites in the GI tract). So on a full
>gluten diet, the totally non-compliant celiac's risk of cancer is an extra
>100% that of those following a gluten free diet. Obviously unacceptable.

Well, yes it is, but the widespread underdiagnosis of cd that is usually
suggested by most researchers in this field would suggest that perhaps the
figures for the incidence of malignancy in the general population are
inflated by the undiagnosed celiacs. So the two-fold relative risk may, be
much greater if endomysium antibody testing is reflective of the true
incidence of celiac disease, at 1:250 in the *healthy* population, and much
higher among those with autoimmune diseases.

>Pick your level of comfort now for whatever cancer risk you're willing to
>live with. 5% of 40 grams of gluten/day is 2 grams/day, for example, for
>that extra 5 % risk. Whether the very ambiguous "malt flavorings" or
>vinegar, there's no 2 grams of gluten in a single portion. So once you do
>that,  in choosing how much more carefully to eliminate gluten or verify
>gluten free, watch instead your nutritional levels regularly: ferritin,
>calcium, sodium, the B vitamins, etc. Monitor antibodies regularly if you
>already know you display them, and biopsy as needed to verify a healthy
>intestine.

That, even with healthy villi, will not alter cancer risk if my hypothesis
is correct (3). I postulate that exogenous opioid peptides from the pepsin
digests of prolamines are absorbed into the blood, and subsequently attach
at the hypothalmic-pituitary-adrenal axis, and thus downregulate activation
of natural killer cells in the same manner that endogenous opioids and
intracranial injections of mophine have been demonstrated to downregulate
natural killer cell activation (4). Natural killer cells, of course, provide
an "immediate first line of defence against cancer cells"(5).

>Cancer, yes, but in proportion as detailed above. Neuropathy, only through
>vitamin deficiency (K and E), which I'm advocating monitoring.

Perhaps deficiency in these vitamins results in some cases of neuropathic
celiac disease. But an autoimmune attack on the myelin sheath is, in my
view, a more likely possibility in many cases of neuropathic celiac disease.

> Another
>autoimmune caused by eating gluten? NO, NO, NO. Yes, the incidence of
>normally rare autoimmune diseases is higher in celiacs than in the general
>population. Genetic connection only: each autoimmune diease has one or
>more pairs of genes plus an environmental stimulus that cause it. All have
>one of these on the same chromosome. Nearby genes tend to stay together in
>the dance of the chromosomes in sexual reproduction. Later generations
>will retain the double pair even if neither disease was expressed
>earlier.

The assumption that autoimmune overrepresentation in celiacs is because of
proximity of chromosomal location  is one perspective. It is equally
possible, and increasingly accepted, that molecular mimicry is the root of
autoimmune diseases in celiac disease.

While I appreciate your well considered commentary on my prior post, and I
grant that your view may be correct, I still disagree with your assessment
of mechanism at work in the increased incidence of malignancy in cd. You
have chosen the conservative report of a 10% incidence, and thus the
conservative view that cd involves only a 100% increase in risk. Taking an
average of the highest and lowest reports I have seen, the increase would be
more in the order of 300%.

I hasten to add that such discussions as these are valuable for provoking
further thought and perhaps, investigation.  While we disagree, I hope we can
continue to exchange ideas in the sharing spirit of this list.

Best Wishes,
Ron Hoggan    Calgary, Alberta, Canada

                             Sources:

1. Fundia, et al "Chromosome instability in untreated celiac disease
patients" _Acta Paediatr Suppl_ 1996; 412:82-84

2. Maclaurin B, Cooke W, Ling N, "Impaired lymphocyte reactivity against
tumour cells in patients with coeliac disease" _Gut_ 1971; 12: 794-800

3. Hoggan R "Considering Wheat, Rye, and Barley Proteins as Aids to
Carcinogens"
_Med Hypot_ 1997; In Press

4. Black, Paul "Psychoneuroimmunology: Brain and Immunology" _Scientific
American: SCIENCE & MEDICINE_ 1995; 2(6): 16-25

5. Tortora & Anagnostakos _Principles of Anatomy and Physiology_ 1990;
Harper & Row, New York.