Friday, October 16, 2009

Why We Can't Study Kibble Effects, When All Dogs Eat Kibble

A friend who read the preceding post said it was (1) boring and (2) confusing. Evidently, I failed to make the scientific points interesting and clear. Let me try again. The key is variation in both predictor and outcome measures.

Let's agree that lack of sufficient amounts of calcium causes rickets, or bone malformations, in developing children.

Let's suppose we study the effects of calcium on rickets in a dairy-rich population. When children are weaned, their diet includes milk, cheese, eggs, and other calcium-rich foods. No rickets is found. We can't study rickets in this population, because all children receive calcium-rich diets and there are no cases of rickets. There is no variation in diet predictors and no variation in the rickets' outcome measure.

Now let's consider the same study in a malnouished population with calcium-poor diets and universal evidence of rickets. Again, the study will fail to show any connection between dietary calcium and rickets, because there is no variation in predictor diets and no variation in rickets' outcomes.

A meaningful study of a causal link between dietary calcium and rickets requres variation in both the predictor diets and outcome rickets. We need a population in which some individuals have calcium-rich diets and some don't and where some children have rickets and some don't. Then we can measure rickets and see if there is an association with amount of calcium in their diets.

That brings me back to kibble. When more than 90% of the pet population is fed exclusively on manufactured foods, it is impossible to show that these foods cause the many chronic dseases pets suffer in the kibble-fed pet population. There is no variation in the predictor diet, so there can be no association with chronic disease outcomes. No one can find the association between commercial pet foods and illness in a pet population that is fed exclusively on kibble.

To study the harmful effects of a kibble diet, we need a pet population in which some individuals are fed varied fresh foods, some are fed raw-meaty-bones, and some are fed kibble diets. Then, we can look for associations between diet and health. Unfortunately, this hypothetical population of pets with a variety of diets does not exist.

Every day in their clinics, veterinarians see cases of chronic diseases, cancers, and immune-system disorders. They don't connect these diseases with kibble diets, because all these dogs, and others in their practices that are not yet ill, are fed kibble. Diet is not a variable, so it is not associated with diseases in their clinic population.

If vets saw a lot of dogs that are fed raw-meaty-bones, they would see shiny teeth, healthy gums, and robust good health. But they don't see raw-fed dogs, because they seldom need veterinary attention.

Bottom line is there is no association between diet and health in populations where more than 90% of pets are fed kibble. Only when altenative diets are sufficiently numerous to create dietary variance will the association of kibble with ill health be found.

No comments:

Post a Comment