Another study was released recently that purports to “prove” that gluten-free diets are associated with increased risk for type 2 diabetes. As with many studies of this type, the findings were misinterpreted but fed into the media’s continual need for titillating headlines. I thought this hubbub would pass by now, but reports about this study (such as this piece of tripe from The Washington Post) seem to be gaining more traction than usual, fueling the misunderstanding and misinformation that plagues nutritional thinking. While I thought this would just pass, it looks like it will not and I’m therefore posting my comments.
First, a few words about epidemiological studies of the sort this group used, the Physicians’ Health Study population of health professionals. The participants were asked diet questions, then health status was tracked over several years. Putting aside the imprecision of such dietary recall questionnaires, we know that such studies simply cannot—no matter how large the study, no matter how meticulous the questions—establish cause-effect relationships; they can only suggest a potential association. The purported 13% difference in type 2 diabetes incidence is minor, given the dramatic imprecision of epidemiological studies; confident associations are typically much larger than this: 40% or 50%, for instance. This does not stop, of course, media people, who are journalists at best, paid marketing people for the grain industry at worst, to propagate their misinterpretations.
To further illustrate the problems inherent in epidemiological studies, let’s pretend that we want to establish whether a Toyota Prius is a safer car to drive than a Corvette. We therefore identify 1000 Prius drivers and 1000 Corvette drivers. We then ask the drivers every year whether they’ve had a car accident. What if we observe that people driving Corvettes are 30% more likely to have a high-speed car accident and 18% more likely to die in the accident? By these observations, should we conclude that Corvettes are far more dangerous cars than Priuses? Of course not. (I made these figures up, of course, just to illustrate, though I wouldn’t be surprised if such numbers approximate the real-life situation.) We can only say that there is an association between driving a Corvette and more accidents and fatalities. But we can most definitely NOT say that the increased dangers are due to the car per se, because it might be due to 1) the kinds of people who buy Corvettes (speed-hungry males), who are more likely to 2) drive aggressively, 3) drink and drive, and 4) ignore laws not to text or engage in other risky activities while driving, etc. You get the picture–it may have nothing to do with the car itself, even though there is an apparent association. But outlets like The Washington Post and others can still make ridiculous headlines claims like “Corvettes found to kill drivers.”
So it goes with trying to track the consequences of consuming more or less gluten. Reduced “gluten” consumption can be a marker for:
- Increased consumption of gluten-free foods made with cornstarch, rice flour, tapioca starch, and potato flour. I have discussed the dangers of these awful gluten-free replacements many times, as they are clearly associated with increased risk for type 2 diabetes, weight gain, inflammatory conditions, heart disease, cancer, and dementia. Sadly, most people who reduce or eliminate gluten gravitate towards such horrible replacement foods that yields an apparent association that cutting gluten increases type 2 diabetes.
- Reduced prebiotic fiber intake–Despite all the toxic components of wheat and grains, there is actually something good in them: prebiotic fibers. The average person obtains around 2-4 grams per day of the prebiotic fibers, arabinoxylan, amylose, and others, from fibers in grains, in addition to inert cellulose fibers. Reduced gluten is therefore a marker for a reduction in prebiotic fiber intake. This is what happens to the majority of people on low-carb, Atkins, and paleo diets if they are not informed about the importance of including plentiful prebiotic fibers in their daily diet. Obviously, you can easily compensate for the removal of grain-sourced prebiotic fibers by increasing consumption of legumes, root vegetables, and other sources. Because the concept of prebiotic fibers is so new, very few people have had a chance to take advantage of this strategy and it therefore will not show up in this epidemiological analysis, instead yielding the apparent low-gluten/more diabetes association.
Those two factors, in addition to other potential unidentified factors that can plague such study design and can only be speculated on, make it impossible to conclude that reduced gluten leads to increased type 2 diabetes or that increased gluten prevents diabetes. Yet that is what was reported.
If we believe the interpretations of the findings offered by the authors of the study, as well as by the generally naive media, then it appears that gluten or some other component of grains are therefore protective against developing type 2 diabetes. Of course, outside of prebiotic fiber intake associated with gluten intake, this is absurd—no such science exists.
The Washington Post, as well as the researchers themselves, were not the only people to misinterpret the findings so tragically. Fox News, ABC News, Boston Magazine and many others made the same absurd blunders. In fact, this silly piece of epidemiological semi-fiction received such favorable treatment in the media that it makes me wonder if there was some sort of financial catalyst, AKA the grain lobby, involved. I don’t know, but it looks and smells an awfully lot like a smoking gun.