Tight control of diabetes? Not so fast…

Link: DB’s Medical Rants � Blog Archive � Tight control of diabetes matters.

OK…several problems with this euphoria concerning an article in the NEJM.  This study was an epidemiologic follow up of the DCCT cohort.  A little background – when the DCCT came out, people went ape about how much tight control of diabetes benefits those with diabetes.  Never mind that the DCCT only included type I diabetics – everyone wanted to apply the results to type 2 diabetes.  EVEN when the UKPDS trial came out – showing that tight control had no effect on patient oriented outcomes of type 2 DM – there was STILL argument.

I’m not usually one to appeal to pathophysiological rationale, but geez…the pathophysiology even suggests that there should be different treatment.  It does injustice to our basic understanding of our bodies’ very complex physiology to say that just because the glucose is elevated, more insulin is needed – when in type 2 DM, there’s already PLENTY of insulin around (doing its best to worsen the situation by causing weight gain and who knows what else…)
Now comes the DCCT followup.  In this post linked to above, rcentor notes jubilantly: "We were right!" and then quickly notes:

One could argue that we only know that this is true for type I
diabetes, and thus we do not have the evidence to push as hard with
type II diabetes. That reasoning is (in my opinion) flawed. (see
previous rant)

What’s it gonna take?  I couldn’t find the "previous rant" he refers to on his site, but I offer the following list of my problems with the whole subject…

  1. There are NO absolute statistics concerning the major endpoints in this article, and no calculated NNTs.  Now call me an EBM wacko if you must, but I am immediately suspicious of enthusiasm about this if I would have to treat hundreds of type I diabetics with labor, time and money intensive treatment to prevent a single cardiovascular event.  Looking at the y-axis of the survival curve, it looks as though the NNTs would indeed be that large.
  2. Whether you’re an empiricist or a pathophysiologist (I know there’s a better term, but I can’t think of it) – there are convincing arguments AGAINST intensive sugar control in type 2 diabetes (in terms of its contribution to decreased mortality).  They’re no more the same disease as diabetes mellitus and insipidus ("but they both produce lots of urine when they’re out of control…they both MUST respond to insulin!")
  3. Intensive control is hard to do.  The NIH is STILL trying desperately to fund projects to figure out how real doctors can do this.  The DCCT follow up itself noted that the vast majority of patients were not continued on intensive control after the study in either group (though intererestingly their HBA1Cs did decline).

The article misleads, and, in my opinion, this blog post’s author is overly optimistic.  There will be harm to intensive control of diabetes – that of opportunity cost (other interventions and health promotions that don’t get done because of the focus on blood sugar).  It does not appear that this study’s outcomes are big enough to offset that harm.