Drinking one or more soft drinks per day, be them diet or regular, may increase your obesity risk. Okay, fine. This is old news by this point. But, for some strange reason, I can't seem to get past this report, which appeared in the very well-respected journal Circulation. This is not because I'm overly concerned that it may be true, but because I feel it's an obvious sign of poor statistical analysis.

What the researchers found is that downing one or more 12-ounce servings of soda per day can increase a person's risk of obesity by 31 percent. Clearly this makes sense when referring to the drinkers of regular sodas, as it is almost common knowledge that soda contains an inordinate amount of fast-digesting, simple sugars and a great deal of calories. My issue, however, is with the claim that diet soda can have the same effect.

There is a well-known Mark Twain quote that, at one point or another, passes through the lips of just about every college statistics professor on the planet. "There are lies, damned lies, and statistics," Twain asserted, pointing to the persuasive power of numbers and their ability to bolster inaccurate arguments. I feel this is undoubtedly the case with this matter involving diet soda and its supposed correlation with obesity. This actually brings me directly to yet another problem I have with this report. Correlation and causality are two very different things; the first of which meaning that A may do something if B does, whereas the latter means that A is the actual cause of B doing whatever it does. They are two very, very different ways to describe the relationship between two variables.

For example, the results of a study on child drownings once revealed that when ice cream sales are up, there are more incidences of kids drowning in pools, lakes, oceans, etc. Does this mean that parents should prevent their children from eating ice cream, for fear of losing them at a young age? The data supported this CORRELATION. However, it did not take long for people to realize the flaw in these research findings: Outside variables had not been controlled for. Essentially, the researchers did not take any other possible factors into account. Obviously, the reason why more children drown when ice cream sales are up has nothing to do with ice cream itself. Rather, it has everything to do with the season in which ice cream is mostly frequently bought and sold. Also known as Summer. Also known as the time when kids spend the most time in and around pools, lakes and the ocean. Also known as the CAUSE of this relationship.

Although I don't know, or even claim to know, for sure that a research flaw of this kind is also responsible for the diet-soda/increased obesity risk study results, I certainly have my suspicions. Did the researchers, just like the researchers behind the ice cream/child drowning study, fail to control for outside variables? And, in doing so, did they base their results on a correlation rather than causation?

I'm inclined to think so. What about you?