The hyperbole occurs when some study shows a small difference in blood glucose control or a small difference in the risk of diabetes when some factor or other is given to a group of people, or just a group of mice.
A good historical example is oat bran. Studies, many of which were sponsored by companies producing oat bran, showed that fiber reduced cholesterol levels. Suddenly, every product on the supermarket shelves, except perhaps toilet paper, trumpeted "Contains oat bran!" as if eating small amounts of oat bran while continuing to shovel in tons of unhealthy pizzas and cheesecakes would make you immune from heart disease.
A current example is a study showing that coffee increases BG levels in people with diabetes. Some people are considering depriving themselves of the pleasure of a cup of coffee or two in the morning. But if you look at the research paper cited by the journalists (which unfortunately, isn't available online unless you pay for it), you'll find that the study subjects had a "standard breakfast" containing 90 g of carbohydrate. This is equivalent to 6 carbohydrate exchanges, a lot even by ADA standards.
After this breakfast, participants who also ingested caffeine went up to about 175 mg/dL; those without caffeine peaked at 171. Big deal! The caffeine drinkers also went up more quickly and came down more slowly, so the total area under the curve was significantly different, after an "ad lib" lunch and dinner as well as the huge breakfast. But they also seemed to be going lower at bedtime (not low; just lower: 117 instead of 126) with the caffeine, and the researchers didn't measure overnight levels.
Cutting back on their carbohydrate intake would have resulted in much larger changes in blood glucose levels than cutting back on their caffeine intake. But this was never suggested. Instead, they concluded, "The presence of hyperglycemic effects in these free-living individuals raises concerns about the potential hazards of caffeinated beverages for patients with type 2 diabetes. Repeated episodes of elevated glucose readings from daily consumption of caffeinated beverages could impair clinical efforts aimed at glucose control and increase the risk of diabetes complications."
Of course the conclusions are what the journalists and popular medical journals will read and trumpet to the world. I can imagine headlines saying things like "Caffeine causes diabetes complications." So some patients reading these stories will stop drinking coffee with their doughnuts, but they'll keep on eating the doughnuts. It's sort of like the people who order a triple cheeseburger and double fries and a diet soda.
Another example of bad information comes from the surprising announcement recently that the ADA now supports low-carb diets for short-term use in weight control.
