B12 for stroke prevention: an exercise in motivated reasoning

Disclosure: I am a member of the Physician Advisory Board and freelance writer for the Medical Post.

I opened my print copy of the Medical Post last month to find this, an article which attempts to make the argument that there is a link between B vitamin deficiency and stroke, and that supplementing with methylcobalamin reduces the risk of stroke.

For anyone familiar with the evidence on this topic, there is absolutely no high-quality evidence to support this claim.

Rather than write a rebuttal, I thought it would actually be clearer to do a line-by-line evaluation of each claim/study cited in the article, and to give the actual context of what the evidence actually shows.

Here we go (the article will be presented in italics):

The first major trial of B vitamins to lower homocysteine for stroke prevention was the Vitamin Intervention for Stroke Prevention (VISP) trial, published in 2004.23 It showed no benefit.

He is absolutely correct. Here’s the link to the JAMA article from 2004. 3,680 patients, multi-centre trial. No effect on vascular outcomes.

Then in 2006 the Norwegian Vitamin Trial (NORVIT) and HOPE 2 trial were published in the New England Journal of Medicine, with an accompanying editorial by Loscalzo, who suggested that the null effects may be due to high levels of unmetabolized folic acid. Following that, the cant became “homocysteine is dead.” 

The NORVIT trial was an RCT 3,749 patients looking at B vitamin supplementation and an outcome of a composite of recurrent myocardial infarction, stroke, and sudden death attributed to coronary artery disease. No benefit.

HOPE 2 randomized 5,522 patients to B vitamins or placebo and found no benefit in the the primary outcome of a composite of death from cardiovascular causes, myocardial infarction, and stroke.

Keep in mind what preceded these studies. Years of observational trials showing that  lower homocysteine levels are associated with lower rates of coronary heart disease and stroke. Folic acid and vitamins B6 and B12 lower homocysteine levels. Thus, one would think that B vitamins should reduce cardiovascular disease. But this is why we do clinical  trials, to see if treatments truly work, not just work hypothetically.

However, the HOPE-2 trial actually showed a 23% reduction of stroke, in a population with better renal function than the other two trials.

In HOPE-2, stroke was not a primary outcome, but was part of a larger composite vascular outcome. There was no reduction in the primary outcome with the intervention of B vitamins. I wish we could ask the HOPE-2 study’s authors whether they think the effect on stroke is meaningful! Oh wait, we can“With regard to the risk of stroke, we observed an absolute reduction of 1.3 percentage points and a relative reduction of 24 percent among patients assigned to the active-treatment group. However, these results must be interpreted with caution. The number of strokes in our study was much lower than the number of coronary events, the confidence intervals around the estimated risk reduction are wide, and the results are not adjusted for the multiplicity of outcomes compared. Also, we found no effect of treatment on transient ischemic attacks. From a biologic perspective, a treatment benefit restricted to stroke would be difficult to explain. Furthermore, the two other large trials of homocysteine-lowering vitamins that have been completed did not show a beneficial effect of treatment on stroke. Therefore, we believe that the apparent beneficial effect of B vitamin supplements on stroke in our trial may represent either an overestimate of the real effect or a spurious result due to the play of chance.” I guess they weren’t impressed either.

An analysis of the VISP trial excluding participants with impaired renal function showed a 34% reduction of stroke/MI/cardiovascular death comparing high-dose vitamins among persons who could absorb B12 well versus low-dose vitamins in persons with lower B12 levels at baseline.

This can be described as “torturing the data until it confesses”. They took the original VISP data, then did a subsequent analysis of whether looking only at normal renal function patients AND excluding those with high and low B12 vitamin levels would result in a lower stroke risk. Doing post-study analyses like this are potentially hypothesis-generating, but cannot be seen as reliable to guide clinical practice.

The French Su.Fol.OM3 trial, in a population with even better renal function and a lower dose of B12 (only 20 mcg daily), reported a 43% reduction of stroke. 

Here again, the primary outcome was major cardiovascular events, defined as a composite of non-fatal myocardial infarction, stroke, or death from cardiovascular disease. No benefit was seen in that primary outcome. Now as for that secondary outcome of stroke, I wonder what the authors had to say….why look at that! They did address it! “In our trial, allocation to B vitamins was associated with a 46% reduction in the risk of ischaemic stroke among patients assigned to B vitamins and a trend towards increased cardiovascular mortality, with an unexpected, significant increase in the number of deaths from any cause among patients receiving B vitamins (not specifically linked to cancer). The latter two results must be interpreted with caution, since the number of statistical tests performed on secondary end points was large and the number of events was small, and the confidence intervals around the relative risk estimates are wide.” I cannot say enough about the humility and academic honesty of the authors of these major trials, and it is truly an example of the peer review process working well. Consider how dangerous it would be for researchers to be claiming that every secondary outcome measure is indeed evidence in favour of treatment. Our clinical lives would be chaos with the degree of medical reversal that would ensue with subsequent trials showing conflicting results.

Then in 2010, we reported significant harm from B vitamins including 1,000 mcg/day of B12 (cyanocobalamin) among patients with diabetic nephropathy. Decline in renal function was accelerated by high-dose vitamins, and cardiovascular events were doubled. In 2011 Stampfer and I hypothesized that harm from cyanocobalamin among participants with renal impairment cancelled out the benefit of B vitamins among participants with good renal function.

The JAMA study from 2010 did indeed show that in patients with diabetic nephropathy, B vitamins worsened both eGFR and vascular outcomes. But to assert that somehow implies that B vitamins would thus reduce vascular risk in those without renal disease? There is simply no logic there. Replace B vitamins with NSAIDs. In patients with diabetic nephropathy, NSAIDs are certainly known to worsen both eGFR and vascular outcomes. Following his logic, in patients without renal disease, NSAIDs would then reduce vascular outcomes? That simply isn’t true, and is in fact quite the opposite.

The final piece of the puzzle fell into place with the publication of the Chinese Stroke Primary Prevention Trial, which reported a 25% reduction of stroke with folic acid alone in patients with hypertension receiving enalapril. Among participants with LDL-C >2 mmol/L, the reduction of stroke was 31%.

The study randomized patients to enalapril plus folic acid or enalapril alone, and did in fact find a reduction in stroke risk. But here were the authors with an explanation of why their findings differed from VISP and NORVIT: “In this population without folic acid fortification, we observed considerable individual variation in plasma folate levels and clearly showed that the beneficial effect appeared to be more pronounced in participants with lower folate levels. In comparison, the VISP study was conducted in the United States, a region with folic acid fortification. Mandatory folic acid fortification in North America has had a significant positive effect on the population’s plasma folate levels. The mean folate levels at baseline in the VISP study was about 28 nmol/L (12.4 ng/mL), which was about 50% higher than that in the CSPPT trial. Therefore, it is not surprising that previous folic acid trials conducted in high folate regions generally yielded null results, which were likely due to the “ceiling effect” of folic acid supplementation.” Pretty clear reasoning. Our rates of folic acid deficiency in North America are so low that in Ontario, the government will not actually pay for the test to be performed, because it is of such low value.

Importantly, among participants with impaired renal function, folic acid slowed the decline of renal function and reduced a composite event that included progression to dialysis and mortality. 

Again, another post-hoc analysis aimed at mining a trial for data. You’ll notice that no nephrologists are actively pushing folic acid on their patients for vascular protection.

This means that the problem with the early trials was not unmetabolized folic acid; it was probably due to cyanocobalamin. A meta-analysis of trials stratified by renal function and dose of cyanocobalamin supports that hypothesis.

They looked at the VISP data again! 13 years later! If at first you don’t succeed….

 B vitamins do reduce the risk of stroke, if not MI, but we should be using methylcobalamin or oxocobalamin, not cyanocobalamin. It is important to use B12 in addition to folic acid because of the very high prevalence of metabolic B12 deficiency, which is largely unrecognized.

If you move the goalposts enough times, eventually the field goal will be good.

Only a small fraction of total serum B12 is active (~ 6-20%). This means that a serum B12 in the lower end of the reference range (~ 160-600 pmol/L) may not contain adequate levels of active B12. In order to confirm adequacy of functional B12 it is necessary for the serum B12 to be above 400 pmol/L, or it is necessary to measure holotranscobalamin, or one of the metabolites that becomes elevated in metabolic B12 deficiency: the specific metabolite, methylmalonic acid (MMA), which is not available in most routine biochemistry labs, or total homocysteine (tHcy), which is a reasonable surrogate for MMA in folate-replete patients. Since we have folate fortification of the grain supply, it is reasonable to measure tHcy for this purpose.

None of this is remotely relevant to clinical outcomes. It’s mechanism-based jargon. Testing total homocysteine for what purpose?

Vitamin B12 deficiency is extremely common in the elderly; ~20% of the elderly have inadequate active B12; among vascular patients the percentage is much higher: in my stroke prevention clinic, metabolic B12 deficiency was present in 10% of patients below age 50, and 30% above age 70.

Why does this matter? Not only does B12 deficiency cause neuropathy, myelopathy and dementia; it also raises levels of tHcy, a clotting factor that markedly increases the risk of stroke. High levels of tHcy quadruple the risk of stroke in atrial fibrillation.

Yes, and nothing we have done in North America to reduce homocysteine in isolation has ever done anything to reduce vascular risk. Time to put this theory to bed. Again. Research dollars are a finite resource, and should not be spent chasing this theory any further.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s