Correlation and causation get confounded because our brains rely on our “vision”, what we see is what we believe.  (BTW, this discussion began yesterday with this post- there will be two more in a row.)  And, most of our causal explanations are naïve (even if they seem complex)- they rely on the fallacy of a single cause (or two).  These oversimplifications help us understand phenomena, so we are positive they are correct.

But, most of what we examine does not have a single cause.  Our body can convert the food we eat to carbon dioxide and water, to simple sugars for metabolism, or to fat.  We each have different metabolic rates- and each do different exercises, which means we can’t predict the pathway a given food digestion will take.

Scientists use statistical correlations to attempt to discover cause and effect relationships.  And, they theoretically understand that correlation is not causation, but if the results follow statistical correlations and are consistent, that line is crossed for them.  Given a “significant” result (one not produced by chance less than 3% or 5% of the time), we conclude that the correlation is not a coincidence.

But, this reliance is no longer a “truism”.  Let’s consider cholesterol levels in our body.  We need cholesterol- but too much of a certain kind leads to blockages in our arteries. The various pathways for it are well-known.  So well known, that a new blockbuster drug- one that was to replace Lipitor (which lowers the level of low-density lipoprotein by blocking a liver enzyme)- was about to be introduced into the marketplace.

This drug, torcerapib, blocked the conversion of high-density lipoprotein into the low-density variety.  Ipso facto; more HDL (‘good’ cholesterol), less LDL (‘bad’ cholesterol). Initial clinical trials proved the facts.  Things were looking really good. But, right before the large scale implementation of phase III clinical trials, the drug was pulled from consideration.  Instead of improving the health of the patients, the drug led to more chest pain, heart failure, and more than a 50% increase in mortality.  This scenario, obviously, was the result of the oversimplification of an issue with multiple causes.  (Now may be a good time to review my discussion of the various cholesterol types- which is way more than just LDL and HDL.)

This is the same reasoning that led to performing operations to fix knees and shoulders that may not need ‘fixing’.  Or, the treatment of back pain (one of the most common ailments)- where just waiting six weeks is often the best therapeutic modality, and not the operation that is often suggested.  All of these operations (surgeries) are costly experiences- in dollars, time, and pain. (You may want to reread my post on the epidemic usage of MRI and reliance on its “results”.  Those images of abnormalities are not the ’cause’ of the pain.)

And, this is the reason why a lot of brain research is being cut back.  This “organ” is way too complicated, comprised of way too many networks, to be assured that a single drug can deal with all the causes to eradicate a single disease.  And, it’s why stem cell research, that provides replacement tissue and organs- that cover the potential gamut of causes- has so much more utility.  And, why we need to know more of the inputs – and sideshows- that affect our businesses.

Share this: