Effect of B-Vitamin Therapy on Progression of Diabetic Nephropathy: A Randomized Controlled Trial

A recent report AGAIN in JAMA (i) has highlighted a potential problem with very high dose oral folate along with B6 and B12 in people with diabetic nephropathy -  “A multicenter, randomized, double-blind, placebo controlled trial (Diabetic Intervention with Vitamins to Improve Nephropathy [DIVINe]) at 5 university medical centers in Canada conducted between May 2001 and July 2007 of 238 participants who had type 1 or 2 diabetes and a clinical diagnosis of diabetic nephropathy.”

Patients in the trial were followed for up to 3 years AFTER the folate and B6 and B12 were stopped.

This trial found a small but statistically significant decrease in GFR in the B vitamin group vs. placebo.  The trial also found a doubling of MI and stroke in patients in the B vitamin group – 3 years AFTER they stopped taking the trial vitamins.

“Our trial is the first study to our knowledge to show significant detrimental effects from pharmacological doses of B vitamins (folic acid, vitamin B6, and vitamin B12).  Other studies have suggested potential harm, or at best a neutral effect of B vitamins. Our results are consistent with a recent small study showing increased vascular events in patients with unstable angina and non–ST-elevation MI, particularly in the subgroup with diabetes.”

It is not reasonable to treat patients for 36 months with a few micronutrients and realistically expect to see any effect on health outcomes.  Especially 3 years AFTER the nutrients were stopped.  The B vitamin group had their homocysteine lowered by only 2.2 micromolar during the trial.  MI and stroke are the result of chronic disease that takes decades to develop, and lowering homocysteine levels by a such a small amount for 3 years out of an entire lifetime is not going to show up as significant differences in cardiovascular outcomes between the groups.

Did the study group the “bad” endpoints and arrange the data so that a statistical effect appears?  Can the stats really be believed?

The only nutrient that showed up in the stats as any sort of potential risk was folic acid.  Trial participants received a placebo (n = 119) OR a daily vitamin pill containing: 2.5 mg folic acid, 25 mg B6, 1mg B12 (n = 119) - that’s a lot of folic acid.  This is a totally unbalanced supplementation regime and there does not appear to be any good reason for choosing such a high dose of folic acid above the other nutrients (or anything else that will lower homocysteine).  Despite this, they have not proved that all this folic acid has done any harm.

One VERY salient point which the examiners appear to have "overlooked" appears in the table of baseline characteristics under “History No%” the number of patients with peripheral vascular disease is SIGNIFICANTLY higher in the B-Vitamin group - 24 (20.2%) compared with the placebo group - 17 (14.3%).  Even a primitive understanding of peripheral vascular disease says that such patients are at greater risk of stroke and cardiovascular disease and decreased kidney function.  Were these the same patients that had MIs and stokes in the treatment group?  Individual patient data are not published so we cannot check this, but this is 7 more people compared with placebo group.

In the placebo group 12 more people had suffered previous heart attacks or angina.  Maybe it is as simple as this – during the trial the treatment group “caught up” to the placebo group as a natural progression of CVD, i.e. in the end the heart attacks etc. were all pretty even over an entire lifetime.

What about this difference in baseline characteristics of the two groups?  Apparently the groups were randomised.  If the baseline characteristics (previous MI/angina and peripheral vascular disease) can be so different in the two groups, then if the groups really were random, this must be due to chance.  The increase in “bad” endpoints (MI, stroke and revascularisation) seen in the Vitamin treated group is very small - in the placebo group (over 3 years) 4 people had a MI and 1 person had a stroke and in the treatment group 8 people had an MI and 6 people had a stroke.  These were the only significant differences statistically between the groups other than the effect on homocystine.  These total numbers are roughly the same as the differences between the groups to start with, so why can’t these be due to chance just like the baseline differences?  Maybe the groups weren’t really correctly randomised to start with? 

That would make the whole trial invalid.  It is a bad study.  It claims a risk derived from very small numbers in groups who were significantly different to start with.

(i)  House AA, Eliasziw M, Cattran DC, Churchill DN, Oliver MJ, Fine A, Dresser GK, Spence JD. Effect of B-vitamin therapy on progression of diabetic nephropathy: a randomized controlled trial. JAMA. 2010 Apr 28;303(16):1603-9. PubMed PMID: 20424250