Another widely accepted dogma that comes from observational studies alone was shown to be just our eyes fooling us when surprisingly this week the American Heart Association released an unusual scientific statement with the conclusion that the evidence does not support gum disease as a risk factor or as a cause of cardiovascular disease. For over 20 years it has been widely believed that periodontal gum disease is a risk factor and a cause of coronary heart disease and stroke. The story of how the dogma that poor oral health became nearly universally accepted as a risk factor of cardiovascular disease is worth looking at because it exposes the risks of accepting data from an observational study as true just because it seems to make sense and conform to what we also see in practice.
Essentially all of the evidence used to support the belief that gum disease is a risk factor or a cause of atherosclerosis was from observational studies. In an observational study it is observed that condition A is present more often in people with condition B than in persons without condition B. Many studies showed that patients who had heart attacks are more likely to have bad oral health than patients who have not had heart attacks. This is far different from saying that gum disease causes heart attacks. In an extensive evaluation of all of the studies showing a relationship between gum disease and cardiovascular disease a panel including both dentists and physicians concluded that the evidence simply does not support the conclusion that there is a causative relationship. The problem appears to be that several other risk factors for cardiovascular disease are also put patients at risk for gum disease. These include tobacco use and low socioeconomic status, as well as age and diabetes mellitus. In a controlled study these confounding variables would be considered and “controlled” for in any analysis. In an observational study this is much more difficult to take into account
The association of periodontal disease and atherosclerosis was so in synch with our bedside observations that it was intuitive to accept the association as dogma. For me at least it never occurred to seriously question the relationship. This was in part because of the widespread acceptance of the test hsCRP (highly sensitive C-reactive protein), a test for low-grade systemic inflammation as an independent risk factor for coronary disease. It was easy to infer that because periodontal disease is a chronic inflammatory condition, can lead to bacteremia, and is a potential cause of systemic inflammatory marker elevation, that is “just made sense” that it is a cardiovascular risk factor.
I hope this 20 year walk down the path of least resistance is one I and others will remember when presented with an observational study purporting to show a relationship. Although I tell patients frequently that just because one factor precedes or coexists with another that it does not automatically follow that one causes the other, I too am obviously guilty of falling into this trap.
Sometimes as a medical community we are criticized for insisting on controlled, randomized, blinded studies to prove efficacy of our treatments, tests and procedures. It can be an expensive, time consuming and sometimes frustratingly tedious process. Still, without solid scientific controlled studies we will be at risk of taking what seems to make sense as factual. Bleeding sick patients was accepted as dogma in centuries prior to use of the scientific method, and we need to beware believing everything we see.