Fluke or Trend? Metformin vs Cancer Part 3

Yesterday, we looked at a prospective cohort study suggesting that use of metformin by diabetics improved cancer mortality.  If you've been keeping count, we're now up to three studies linking metformin use to lower risk of cancer and/or better outcomes.  But is that enough to make metformin part of every cancer survivor's regimen?  Perhaps we just happen to have 3 coincidental studies.  And remember that correlation doesn't prove causation.

Thinking along those lines, perhaps I've stayed up too often watching late night infomercials only to remember that famous catch phrase, But Wait, There's More! when I stumbled upon a population-based cohort study published in the February 2006 issue of Diabetes Care in which the authors concluded that diabetics treated w/either insulin or sulfonylureas had greater cancer mortality than those treated w/metformin.  

To arrive at their conclusion, the authors followed for 5+yrs 10,309 diabetics avg 63yo newly prescribed either metformin, sulfonylureas or insulin.  Cancer mortality was 5.8% for insulin users, 4.9% for sulfonylurea users and 3.5% for metformin users.  But recall & apply Einstein's theory of relativity: we don't know whether insulin & sulfonylureas are more dangerous or whether metformin is protective.  All we can conclude is that metformin is better relative to insulin & sulfonylureas when it comes to cancer deaths.

But the good news is that we have the foundation for a trend of benefit from metformin that is less likely to be a fluke given 4 disparate studies.  Now all we need is a randomized, double-blinded, placebo-controlled trial of metformin vs cancer risk & outcomes . . .

Follow @alvinblin
Tweet