Sorting out Cause and Effect
We live in strange times. Though whole genomes, of an individual as well as a malignant progenitor stem cell responsible for AML, have been sequenced and though modern medical diagnositic machinery and implantable devices increasingly look like things that Gene Roddenberry would have rejected as too futuristic, the fact is that scientists know very little about how our bodies work. Take C-reactive protein for example.
Let's say you decide to study heart attacks by measuring the levels of certain proteins in the blood and then later checking to see if there's been an association between the levels measured and an adverse event like a subsequent heart attack. Bingo! People with high levels of C-reactive protein have a high risk of heart attack. Wow!
Ok, let's assume the association is strong and consistent so that the finding probably reveals something about how the body works. But what?
Too often these sorts of findings have precipitated a Whac-A-Mole reaction from the medical community whereby drugs are promptly designed and thereafter prescribed to begin whacking the protein back down to "normal" levels. Causes precede effects, right? So this must be a cause, right?
Well, what if C-reactive protein is actually part of the body's response to accumulating plaque and inflammation? What if its level correlates with risk because it's part of the body's attempt to fix whatever is broken? What if it's not a cause of heart disease but is really an effect of heart disease? Reuters is reporting on a new analysis published in Lancet indicating that at least in the case of C-reactive protein earlier assumptions had indeed confused cause and effect. The authors conclude that it is unlikely that C-reactive protein is a cause of heart disease.
