Motivation behind the research
Dr. Kilmer McCully first reported in 1969 that very high blood levels of an amino acid called homocysteine due to a hereditary disease called homocystinuria caused blood clotting and strokes in young people. Then it became apparent in the 1990’s that lesser elevations of homocysteine also increased the risk of strokes and heart attacks.
Homocysteine is an amino acid that causes increased clotting of the blood and damages the artery lining. About 20% of the population have high blood levels of homocysteine, but it is more common in older persons and those with kidney failure. Blood levels of total homocysteine (tHcy) can be lowered with B vitamins: folic acid, B6 and B12.
Clinical trials of the B vitamins began in the late 1990s. All the early trials used high doses of cyanocobalamin (the form of B12 that has cyanide in it). I was a member of the executive of the Vitamin Intervention for Stroke Prevention (VISP) trial, published in 2004. We did not find a significant reduction of strokes or heart attacks. Then in 2006, two additional large trials were published in the same issue of the New England Journal of Medicine. The Norwegian Vitamin trial (NORVIT) showed no benefit, and in fact showed harm in the arm of the study that included cyanocobalamin. The Heart Outcomes Prevention Evaluation (HOPE) 2 trial actually showed a 25% reduction of stroke, but (it is now clear) was misinterpreted by the authors.
Being cardiologists, they couldn’t think of a biological difference between heart attacks and strokes, and since heart attacks were not reduced, they concluded that the reduction of strokes was a chance finding. In the same issue of the New England Journal, an editorial by Dr. Joseph Loscalzo, then Editor-in-Chief of an important American Heart Association Journal called Circulation, concluded that it was clear that B vitamins did not reduce vascular risk, and that harm from folic acid may have been the reason. Since 2006, most doctors have thought that B vitamins to lower homocysteine do not prevent stroke.
The evidence that homocysteine causes stroke is so compelling that I suspected there was something wrong with the study results. There were several reasons why the VISP trial may not have shown benefit:
- We did not use a placebo control; in order to try to limit the risk of patients going out and buying vitamins on their own, we used a low-dose vitamin versus a high-dose combination. For ethical reasons, the low-dose vitamin contained the usual recommended daily intake of B12 (6 micrograms).
- Folic acid fortification of the grain supply in North America became mandatory just as VISP was getting started, thereby negating the benefit of the folic acid in the combination.
- For ethical reasons, we decided that anyone with low serum B12 levels should be treated with monthly B12 injections, regardless of whether they were randomized to low-dose or high-dose vitamins. This negated the benefit of the vitamins in the very persons who stood to benefit most.
In collaboration with others, I initiated a study to see what happened among the study participants who did not receive B12 shots. For the wrong reason, I also decided we should exclude patients with renal failure. (The reason was that, based on a study I had done in patients with kidney failure, I thought they would not benefit; I never dreamed they would be harmed.) What we found was a 34% reduction of stroke, vascular death and heart attacks, comparing patients who could absorb B12 well and received high-dose vitamins, versus those with poor B12 absorption receiving low-dose vitamins.
In 2010, the French Su.Fol.OM3 study, in younger patients with better kidney function than the earlier studies, and using a much lower dose of cyanocobalamin (20 micrograms instead of 400-1000 mcg daily) showed a 43% reduction of stroke. However, the report of the study focused on the failure to show a reduction of heart attacks, and the stroke reduction received little attention.
Also in 2010, I published with my co-investigators in the Journal of the American Medical Association (JAMA) a study in patients with diabetic kidney disease, reporting that high-dose B vitamins including 1000 mcg daiy of cyanocobalamin were actually harmful: they made the kidney failure get worse faster, and doubled the rate of adverse outcomes (progresson to dialysis and a composite of heart attacks, stroke, need for surgery, and death). In 2011, I hypothesized in JAMA, with Dr. Meir Stampfer of the Harvard School of Public Health, that in the early trials, harm from cyanocobalamin among study participants with renal failure may have cancelled out the benefit of B vitamins in participants with good kidney function.
Then in 2015 the Chinese Stroke Primary Prevention Trial (CSPPT) showed that folic acid alone, in a country where folic acid fortification of the grain supply has not been implemented, reduced stroke by 25%. Among participants with a higher blood cholesterol level (and therefore at higher risk) the reduction of stroke was 30%; importantly they also found that folic acid was beneficial among the study participants with poor renal function. That was the final piece of the puzzle to fall into place, and led to the study we carried out.
With Dr. Graeme Hankey of Perth, Australia, who was the Principal Investigator of a large Australian trial of B vitamins for stroke prevention (VITATOPS), and his principal statistician Dr. Qilong Yi (now in Ottawa, Canada), we performed a meta-analysis of all the B vitamin trials. (Combining results from all published studies is called “meta-analysis”.)
Other meta-analyses had been performed previously, some concluding that B vitamins reduced stroke, some concluding that they didn’t. What was different about our study was that we included the CSPPT results, and, based on all I have told you above, we separated the studies into groups: studies that had poor kidney function versus those with good or unspecified kidney function, and studies that used high-dose cyanocobalamin (400 to 1000 mcg daily) versus no cyanocobalamin or a very low dose (20 mcg daily). What emerged was that in studies with good kidney function and in studies with no/low cyanocobalamin there was a significant reduction of stroke; in studies with poor kidney function and in studies with high-dose cyanocobalamin there was no benefit, and if anything, harm.
This means that B vitamins do reduce the risk of stroke, but there is a catch; B12 with cyanide is harmful in patients with poor kidney function (which includes the elderly).
High levels of tHcy quadruple the risk of stroke in atrial fibrillation, a heart rhythm disturbance that causes clots to form in the upper chamber of the heart on the left (the left atrium). Atrial fibrillation, high homocysteine levels and B12 deficiency are all much more common in the elderly.
Because metabolic B12 deficiency is so common (10% of stroke patients below age 50 and 30% above age 70), and because B12 deficiency not only increases the risk of stroke, but also dementia, it is very important that it be diagnosed. A serum B12 in the normal range (160-600 pmol/L) does not establish the adequacy of functional B12; in patients with a serum B12 it is necessary to assess functional B12 by measuring holotranscobalamin (active B12), or one of the metabolites that become elevated in metabolic B12 deficiency: methylmalonic acid or tHcy.
In elderly persons and those with poor kidney function, instead of cyanocobalamin we should be using a form of B12 that does not contain cyanide (methylcobalamin or hydroxocobalamin).
It is not yet clear how cyanocobalamin is harmful in persons with renal failure. Dialysis patients have high levels of cyanide, which is normally eliminated as thiocyanate (consuming beneficial hydrogen sulfide in its formation). However, the amount of cyanide in 1000 mcg of cyanocobalamin is only ~ 20mcg, whereas normal daily intake of cyanide from foods and smoking is around 150-200 mcg, and not all the cyanocobalamin would be absorbed, so it seems unlikely that the cyanide per se in the cyanocobalamin is the problem. To become active methylcobalamin, cyanocobalamin has to have its cyanide removed; this is called “decyanation”.
It is possible that patients with renal failure, having high cyanide levels, may have blocked decyanation. How this would be harmful needs to be worked out. Furthermore, clinical trials using methylcobalamin or hydroxocobalamin to reduce stroke risk have not yet been carried out, and need to be done. In the meantime, it would be prudent (particularly in persons with impaired kidney function, including the eldelry), to use methylcobalamin or hydroxocobalamin for safety.
B vitamins to reduce the risk of stroke should be used routinely, and all stroke patients and patients with atrial fibrillation should have their serum B12 and tHcy (homocysteine) measured. Metabolic B12 deficiency is common, particularly in the elderly, and usually undiagnosed. This needs to change. We will continue trying to inform physicians about this. Since 1977 I have given 600 lectures in 39 countries, and in recent years a lot of those talks have been about this issue, but most physicians still think “homocysteine is dead”.
Research Article: B vitamins in stroke prevention: time to reconsider. Lancet Neurology, 2017
Call to action
- Most physicians are unaware of this work. Please print this along with the published paper and give it to your personal physician, and ask to have your serum B12 measured. (This is particularly important if you are elderly or have a family history of B12 deficiency or “pernicious anemia”; it can run in families.) If your serum B12 level is below 400 pmol/L, ask to have your holotranscobalamin or methylmalonic acid or tHcy measured. If you have metabolic B12 deficiency and you are elderly or have poor kidney function, take methylcobalamin or hydroxocoalamin (one or the other, but not both, are available in some countries).
- If you would like to support our work at the Stroke Prevention & Atherosclerosis Research Centre, please go to our website and follow the instructions at http://www.imaging.robarts.ca/SPARC/Donations. While there, you can also download the recipe booklet I give to my patients and download some of our other publications.