Dec 15 2016

Abusing Science You Don’t Understand

insulinInterpreting scientific studies is difficult. You need to have a thorough working knowledge of scientific methodology, statistical analysis, and the specific field of the science itself. Individual studies also need to be put into the context of all the other studies that touch upon the same question.

It also has to be recognized that all scientific studies are imperfect and only look at a slice of a larger question. Even scientific experts can therefore honestly disagree about how best to interpret the evidence, until that evidence becomes overwhelming (and even then there are often outliers with minority opinions).

If you have an agenda other than understanding the best interpretation of all available evidence, it is easy to find evidence you can twist to serve your preferred narrative. You can then (falsely) claim that your position is supported by science. You can also rely on the fact that most of the public will not be sufficiently scientifically savvy to see the flaw in your reasoning.

A recent article by Edward Morgan on the alleged risks of GMO insulin is a great example of this phenomenon. The title is: GMO Insulin Causes Type 1 Diabetes in Type 2 Diabetics, Study Finds. That is not what the study found.

The author has a clear anti-GMO agenda and so they are trying to argue that insulin which is made from genetically modified bacteria (which is how almost all insulin is made today) is dangerous. They do this by cherry picking studies they clearly don’t understand and then misinterpreting them.

The first study, which is not the one referred to in the headline, is a 2013 study looking at various outcomes in Type II diabetics based on the treatment they are receiving:

Patients comprised 84 622 primary care patients with T2DM treated with one of five glucose-lowering regimens: metformin monotherapy, sulfonylurea monotherapy, insulin monotherapy, metformin plus sulfonylurea combination therapy, and insulin plus metformin combination therapy. There were 105 123 exposure periods.

They found that a number of negative outcomes, including all-cause mortality, was highest in the insulin monotherapy group. Morgan reported that this study showed that insulin therapy caused these negative outcomes. Again, that is not what the study showed, and is a great example of why you need to understand clinical trial design before you can have any idea what the results mean.

Morgan should have had a clue from the last line of the conclusion:

Differences in baseline characteristics between treatment groups should be considered when interpreting these results.

This was a retrospective study, meaning that the researchers were looking back at medical records to see what had already happened. They were not randomizing patients to different treatment groups and then seeing how they do. This is a critical difference, for the very reasons the authors state. It is possible that patients who receive insulin monotherapy are not the same at baseline as those who receive metformin. There is good reason to think this is the case, as in Type II diabetes patients who receive insulin are generally sicker and have worse diabetes control overall, otherwise they would not have resorted to insulin. If they could be controlled with metformin alone, which is a oral medication, they would have.

Type II diabetics have insulin resistance. They can make insulin (unlike Type I diabetic who don’t make insulin) but their cells do not respond to insulin normally. This leads to hyperinsulinemia, which has its own negative outcomes. Taking insulin is therefore a last resort, and used only when the blood sugar is dangerously high. So of course patients with Type II diabetes taking insulin had worse outcomes than those managed with oral medication.

It is also quite possible that Metformin has some protective effects, and if you are taking insulin alone you are missing out on those protective effects.

The bottom line is that you cannot conclude from this uncontrolled retrospective study that insulin is causing any of the negative outcomes measured. You may also notice that the words GMO never appear in this study, and there was no comparison between GMO-derived and animal-derived insulin.

The more recent study is a small study from Japan, involving only 6 patients, titled: Insulin Administration May Trigger Type 1 Diabetes in Japanese Type 2 Diabetes Patients With Type 1 Diabetes High-Risk HLA Class II and the Insulin Gene VNTR Genotype.

Again you will notice that there is nothing in this study about GMO derived insulin. The purpose of this study was to look for genetic characteristics that might help predict which patients will develop an immune reaction to exogenous insulin. In some, but not all, of the patients studied they found they had a genetic predisposition to type I DM. If these findings hold up with more data it may provide a way of identifying patients who should not receive insulin, or for whom insulin use should be minimized, because they have a genetic predisposition to develop an immune reaction that will result in type I DM.

This is a complicated area of genetics and diabetes. The two genetic variants looked at in the study exist in a minority of patients, especially if you consider those who have both genotypes. Also having both genotypes magnifies the risk of Type 1 DM. So, if this study holds up to replication and further investigation, it seems that having a double dose of genetic predisposition to DM may also confer a higher risk that insulin therapy can trigger an antibody reaction that leads to Type 1 DM.

This says nothing about the majority of patients who do not have this genetic predisposition, and it saying absolutely nothing about GM insulin vs animal-derived insulin. Yet, somehow Morgan extracted from this study an anti-GMO narrative. He states that narrative specifically in his conclusion:

Also, diet is the #1 factor in the pathogenesis of most chronic conditions that afflict the modern world; more specifically, the consumption of foods or food-like products that deviate from our ancestral diets generate the physiological conditions that produce disease in the first place.

There is no evidence to support this position. Of course, diet is extremely important in Type 2 DM. That is why such patient are placed on a diabetic diet, and encouraged to exercise and lose weight. In many patients Type 2 DM can be reversed with diet and exercise. But Morgan is not referring to the science-based approach to DM, but rather to “magical” natural foods and an appeal-to-nature fallacy that somehow “regenerates” tissues and cures diseases.

This is all nonsense, and the evidence he cites does not support his position. The only thing we can conclude from his article is that he is not able to understand the scientific articles he is citing, or he does not care to really understand them. They are a rhetorical device only.

11 responses so far