Statistical Manipulation

You've probably heard that taking aspirin regularly is good for your heart. Not anymore. According to yet another new study, it's now bad for your heart and causes high blood pressure. Ditto ibuprofen, acetaminophen, and pretty much every thing else in your medicine cabinet.

Uh-huh. Sure.

A quick lesson in bad science and statistical manipulation:

Men who took such drugs for most days in a week were about one-third more likely to be diagnosed with high blood pressure than men not taking them, the researchers found.

According to WebMD, 28.7% of the US population had hypertension in a 1999-2000 survey. That's only 4.3 normalized variance points from the claimed one third of newly diagnosed hypertension patients who "took such drugs for most days in a week."

In other words, the statistical diagnostic base of the study matches the non-correlated base for new hypertension cases within a negligible margin of error. When distributed across the entire population, that 1/3 number is marginalized to insignificance in relation to total cases of hypertension vis-a-vis total heavy users of OTC painkillers.

But hey: it got their names in the news, yes? That seems to be the entire point of many "studies" nowadays.