You’ve probably heard the news by now. There’s a new risk factor for heart attack. Move over cigarette smoking, high blood pressure, obesity, and diabetes, and make way for… car ownership.
That’s right. In a study published in the January 11 issue of the European Heart Journal, researchers found that people who owned a car [and a television set] were 27% more likely to have a heart attack that those who didn’t. Surprisingly, this hasn’t led to a sudden spike in rental car sales or a flood of title transfers.
In fact, despite the correlation found in the study, neither the authors of the study nor anyone in the popular media who’ve reported on it have even hinted that owning a car causes heart attacks. Nor have they suggested that the correlation has to do with prolonged steering wheel clutching, toxic car fabrics, city dwelling, asphalt inhalation, road rage, or any of an almost limitless number of potential explanations. Rather, they’ve suggested instead that the real culprit behind the risk increase is something only tangentially associated with car ownership: sedentary behavior. Sedentary behavior, which has nothing at all to do directly with the act of buying or driving a car, is postulated as the hidden variable causing heart attacks in car owners.
Now let’s contrast this with how a recently published study demonstrating a correlation between red meat consumption and stroke risk is reported (we could choose just about any dietary study to make this point). Here we have the exact same type of data – folks who ate more red meat had a slightly higher incidence of stroke, just as folks who had heart attacks were more likely to own a car. In the study on stroke and red meat, however, there’s no mention of any potentially confounding hidden variables. This time, it isn’t something tangentially associated with red meat consumption that’s postulated to account for the correlation. No, despite the fact that once again we could propose an almost limitless number of potential explanations for this correlation, this time we’re led to believe that there’s a direct link between the two, and that these findings “support current recommendations to limit how much red meat people eat.”
To reiterate, when the correlation is between car/TV ownership and heart attacks, the notion of a causal link between the two factors is too absurd to even consider. In the study on red meat and cancer, however, the possibility that there could even be confounding variables isn’t even mentioned.
The truth is, neither of these studies – particularly the conclusions being drawn from them – are examples of good science in action. This type of data should never be used as a basis for health advice, nor does it ever warrant a headline. It represents the lowest quality of scientific evidence – hardly a source for basic truth. At best, correlations like these can only suggest a hypothesis that can be tested further in a more rigorous fashion. The inherent plausibility of any particular hypothesis is simply a reflection of the biases of those interpreting the data.
Yet, this is the type of low quality data, and this is the type of sloppy, careless reasoning that has led to the current prevailing wisdom about proper nutrition. This same error of causal attribution - one that is immediately rejected by all sentient beings as absurd in the study on heart attacks and car ownership – lies at the heart of why we’re told that minimizing animal fat and cholesterol intake is one of the keys to optimum health.
So just what happens when you do make public health recommendations based on low quality data and sloppy reasoning? This:
And in case you’re wondering if this is just because folks aren’t heeding this brilliant advice: