We studied B12 levels in our metformin-treated participants at two time points during DPP. Unfortunately we didn’t have samples available at the very beginning, but we were able to measure samples after people had been treated for roughly 5 years and then again after 9 years of treatment, and we found that there was a significant prevalence of biochemical B12 deficiency that was higher in the metformin-treated participants than in the placebo group. It depends on how you define B12 deficiency, which is sort of an issue unto itself, but the people who definitely had low B12 was approximately 5 percent after 5 years of treatment and went up to about 9 percent after longer duration treatment.
The mechanism for low B12 with metformin, I don’t think it has been completely sorted out, but it appears to be a problem with absorption. It’s not related to intrinsic factors-so the classic pernicious anemia that we all learned about in medical school, that’s not really the mechanism, but there seems to be something about the effect of metformin in the gut that impairs absorption.
What should we do clinically at this point given what we know so far about the relationship between metformin and B12 deficiency? I think there is some debate about whether everyone taking metformin should be treated. There are jokes about should we just add B12 to the metformin tablet to take care of this problem. I don’t think we’re quite there yet, in part because we don’t even know what the appropriate repletion method would be. So, we identified people who have low B12; they’re taking metformin. We think it’s likely that oral supplementation will be adequate, but that really hasn’t been studied in any detail. I think being aware of the possibility of low B12 is important in patients who are being treated with metformin, and ADA guidelines came out this year for the first time recommending periodic testing of B12 levels.
One thing I will mention about our findings in DPP that’s relevant to the issue about how we deal with this clinically is that, from the beginning of the DPP study, we were aware of the relationship between metformin and low B12, so we were monitoring CBCs during the study. But we found in our analysis of the B12 levels that the presence of anemia did not discriminate between who did or did not have low B12 levels, so I think we may have a false sense of security by just analyzing hemoglobin and hematocrit, for example, in our metformin-treated patients. I think that’s one of the reasons why guidelines are now starting to come out recommending actual measurement of vitamin B12.
I think it’s hard at this point to have a precise recommendation for when to begin measuring, but I think within a few years of starting metformin is probably when I would start to consider doing it.
The question clinically is whether we should be treating all patients who are taking metformin with a vitamin B12 supplement, or whether we should be testing them. What’s the most cost-effective approach? I think that one could make an argument for providing B12 supplements to everyone who is taking metformin, but the one caveat to that is that I don’t think that the appropriate treatment regimen has really been clearly identified.
We believe and we expect that higher doses of vitamin B12 will overcome the malabsorption that’s associated with metformin, but we don’t know that for sure, and I think it would be somewhat risky to recommend routine supplementation without some measurement of vitamin B12 levels at some point-just to assure that if there was malabsorption that it had been overcome.
The question can come up whether the risk of vitamin B12 deficiency might be a factor in the decision to actually prescribe metformin. Should this be a reason to avoid prescribing metformin, especially these days when there are many other oral medications or easily administered medications that we could choose from? My personal view is that that’s probably not warranted. There are so many benefits to metformin-such an excellent safety profile and inexpensive, readily affordable, readily available drug-that I think that it can be safely prescribed as long as the potential for B12 deficiency is recognized and monitored.