Correlation Does Not Equal Causation

Science can be a funny thing. And obesity science is generally ambiguous at best. But media reports on science are often downright misleading. Granted, journalists need catchy headlines like Obesity More Dangerous Than Terrorism to draw in readers who are always hungry to hear about the next crisis. But an article that sounds extremely captivating may also be just plain deceitful.

Take the demonization of saturated fat and cholesterol. Many of you are familiar with the flawed science behind the era of hate for eggs and butter. (If not, check out Tom Naughton's video.) In fact, though further research continues to exonerate saturated fat and cholesterol (and even demonstrate their benefits), the majority of Americans are still convinced that egg yolks and whole milk cause disease and obesity.

So it helps to approach scientific research in general with an eye for detail. For example, it helps to understand the fact that correlation does not equal causation. This means that factors can be associated with each other without one thing causing the other. Confusing correlation with causation produces not only flawed reports on health issues, but also reinforces stereotypes and prejudices.


"The majority of knowledge regarding the relationship between health and weight is drawn from epidemiological research. Epidemiological obesity research compares groups of overweight and obese individuals with a control group of normal weight individuals. It is intended to uncover associations which then need further examination. It cannot tell us whether a variable causes or even influences another.

Consider this: It is well established through epidemiological research that bald men have a higher incidence of heart disease than men with a full head of hair. However, this doesn't mean that baldness promotes heart disease or that hair protects against heart disease. Nor is it recommended that bald men try to grow hair or buy toupees in order to lessen their disease risk."

- from page 129 in Health at Every Size by Linda Bacon

"Correlation does not equal causation. There's an example often used in Sociology 101 classes to illustrate this point: In the summer months, both ice cream sales and murders increase, meaning there's a correlation between ice cream sales and murder. Now, in theory, once a correlation has been established, it's worth looking into whether one thing causes the other. Does ice cream trigger homicidal impulses? Do murderers like to celebrate a job well done with a nice waffle cone? No, probably not. In fact, heat is what causes the rise in both ice cream sales and homicide. Correlation does not equal causation."

- from page 175 in Lessons from the Fat-O-Sphere by Kate Harding and Marianne Kirby

So when I bring up touchy subjects as I did when I questioned the obesity epidemic and asked if obesity really killed, I'm not just trying to rock the boat. I'm actually trying to do some critical thinking about the ideas that many of us take for granted because of the attention they receive in the media.

Don't be convinced by reports that use melodramatic language and scare tactics to influence your opinions. Instead, simply be willing to question what you hear and form your own ideas rather than being tossed from one dogmatic regime to another.

I don't know about you, but after spending a few years being seriously afraid of real food like cheese, butter and eggs because a pediatrician told me saturated fat causes cancer, I'm willing to take a step back and give it some critical thought before jumping on the next health bandwagon.

This post is part of Real Food Wednesday.







-----------------------------------
Follow Me on Pinterest