B Vitamin Research Assaulted by Defective Study

I n the February issue of this publication (see “B Vitamins Calm the Seas of Memory”), we reported that several B vitamins (folate, Vitamin B2, vitamin B6, and Vitamin B12) increase cognitive function across a wide range of activities. Also, that “low levels of these B vitamins have been associated with increased homocysteine, an amino acid known to have a direct neurotoxic effect in the brain where it is thought to potentiate beta-amyloid neurotoxicity, and thus exacerbate Alzheimer’s disease.”

However, a new meta-analysis published in the American Journal of Clinical Nutrition by Oxford University researchers contradicts the above with the conclusion that “Homocysteine lowering by using B vitamins had no significant effect on individual cognitive domains or global cognitive function or on cognitive aging.”1

Inaccurate and Misleading

There is significant criticism of the conclusion that B vitamins are without value for brain health. Clinicians and scientists have labeled the study “inaccurate and misleading,” articulating concerns that the meta-analysis could bias research funding and health policy decisions. Also, it could have a negative impact on individuals in the very early stages of dementia — who could miss out on a potentially effective treatment.

According to Dr. Peter Garrard of the Cardiovascular and Cell Sciences Research Institute at St George's, University of London, the analysis of previous clinical trial data cast no doubt whatever on the potential of folic acid and vitamin B12 to prevent dementia, and that the lead author's comments were “unjustified and misleading.”2 Dr. Garrard specializes in neurological disorders, particularly neurodegenerative dementia, cognitive disorders, progressive language disorders, early onset dementia and frontotemporal dementia.

Dr. Garrard incisively said that taking B vitamins lowers blood levels of a homocysteine, which in high concentrations acts as a potent risk factor for dementia. Furthermore, Dr. Garrard emphasized, the “first-rate scientific evidence that the use of B vitamins confers both biological and neuropsychological benefits” on individuals aged over 70, who had noticed a recent decline in their cognitive abilities.

Continuing, he emphasized the urgent need for a definitive trial to establish whether this simple and safe treatment can slow cognitive deterioration in a similar group of people, as such individuals are known to have a heightened risk of developing full-blown Alzheimer’s disease.

Wrong Data, Wrong Test, No Decline, Few Conclusions

Both Garrard and Dr. David Smith, founder director of the Oxford Project to Investigate Memory and Ageing, sent separate letters to the American Journal of Clinical Nutrition, pointing out substantial flaws in the analysis. These include:

  1. Reliance on data from trials of vascular disease prevention rather than dementia

  2. The use of the Mini Mental State Examination (MMSE), which is designed to detect dementia but is unsuitable for assessing small changes in cognitively normal people

  3. The absence of any cognitive decline in untreated patients, rendering the whole study irrelevant to the question of clinical benefits in cognitive impairment or dementia

According to Garrard’s letter, “Although the sheer volume of data incorporated into this analysis testifies to the industriousness of its authors, few other conclusions can be drawn.”3

Babies and Bathwater

“Methodologic shortcomings of individual studies cannot be assumed to have been neutralized by the simple expedient of pooling on this scale,” wrote Garrard, et al. It is essential that all relevant differences between the constituent studies be pointed out. That’s because “the danger of not doing so is that any study giving a strong and interesting signal may be considered unimportant, no matter how well designed and conducted, on the grounds that it bucks a predominantly negative trend: an all too familiar tale of babies and bathwater.”

None of the trials subsumed by the
meta-analysis were designed to
evaluate the effects of
treatment on cognition.

The article in the American Journal of Clinical Nutrition contains at least 2 egregious examples of this practice (see point 2 above). Although MMSE is undoubtedly useful as markers of dementia severity, its summary measures are not insensitive to mild cognitive change. MMSE is also remarkably stable over time in populations of healthy elderly individuals. This means that as indexes of cognitive aging they have little practical use. Also, none of the trials subsumed by the meta-analysis were designed to evaluate the effects of treatment on cognition.

Highly Tendentious Statement

“Finally, we [Garrard, et al] would question the lead author’s highly tendentious statement that the findings of this study of cognitive aging show that ‘‘taking folic acid and vitamin B12 is sadly not going to prevent Alzheimer’s disease,’’ a quotation that was picked up by many of the new and mainstream media outlets that reported on the meta-analysis (see http://tinyurl.com/kb88cvg for the full text of the press release).

Once again, as per Dr. Garrard, et al, there is outstanding scientific evidence that homocysteine reduction with the use of B vitamins confers both biological and neuropsychological benefits on individuals with the clinical syndrome of mild cognitive impairment.

The Tyranny of Very Large Numbers

“We are deeply concerned with the impact that unjustified claims may have on individual members of the public and on policymaking and research funding organizations who, in the absence of opportunities to examine the details of the constituent studies, may allow themselves to fall under the tyranny of very large numbers.”

Bad Choice of Trials, Assessment Tools, and Analysis

Regarding Professor Smith’s letter to the American Journal of Clinical Nutrition:4 “We [Smith, et al] are concerned about 3 aspects of this analysis: the choice of trials, the cognitive assessment tools, and the analysis and interpretation of data.

“This meta-analysis would have benefited from having a clear hypothesis: did the authors wish to examine whether B vitamins enhance cognition? If so, no cognitive change in the placebo group is necessary.

“Or was the hypothesis that vitamins slow age-related cognitive decline, as implied by the authors’ term “cognitive aging”? If the latter, the meta-analysis should have been confined to trials in which the placebo group experienced cognitive decline.

“Unfortunately, most of the original trial reports reveal that data on cognitive change are either not available or that cognitive function did not change.” For 8 of the 11 trials, the placebo group did not show significant cognitive decline, the significance of any apparent decline was not reported, or the original article did not report cognitive test results at all.

“The shortcomings in trial design and flaws in the original reports suggest that these trials should have been excluded from the meta-analysis,” wrote Smith, et al.

The MMSE is a screening test for
dementia but is not sensitive for
cognitive aging.

According to Smith, et al in their critical letter, their second concern involves the cognitive assessment test used. Only 4 trials used domain-specific cognitive tests that are sensitive and appropriate for the assessment of cognitive change, they point out. The remaining 7 trials (containing most of the subjects) used the MMSE. The MMSE is a screening test for dementia and is not sensitive for cognitive aging.

Doomed to Fail

As an example, the letter of criticism continues, at the relatively low mean age of 66 years in the meta-analysis, most will achieve a perfect maximum score of 30, indicating normal cognitive function. Hence, the individual trials using the MMSE were doomed to fail. Such a flawed design in the individual trials is unlikely to be compensated by including them in a meta-analysis.

Their third concern is the analyses and interpretation of the results. For instance, the authors did not identify subgroups and risk groups in each individual trial that, on a priori grounds, were likely to benefit from B vitamins. Instead, the authors combined data for all of the MMSE-type trials, with the result that no significant subgroup effects were found.

Benefits in Several Cognitive Domains

Subgroup analysis should also have been conducted on the individual domain-type trials. Thus, in one segment of the trials, subjects were recruited on the basis of elevated homocysteine, a subgroup likely to benefit. Indeed, the subjects who received folic acid for 3 years showed benefits in several cognitive domains, and therefore it is surprising that the meta-analysis considered the results of the folic acid trial as due to “chance,” given that this study had a better design for testing cognitive change than most of the other trials.

Also, in another of the chosen trials, the participants with low vitamin intake at baseline had slower cognitive decline on B-vitamin treatment, whereas there was no effect in participants with adequate intakes.

Interfering Drugs

Another problem is that other treatments given to the participants might have modified the effect of B vitamins on cognition. In a large trial subsumed by the meta-analysis, both “placebo” and B-vitamin groups received simvastatin, a statin that may protect against dementia, or cause it!

Also, in several of the B vitamin trials the results may have been changed by concomitant aspirin use, which appears to negate the beneficial effect of B vitamins on cardiovascular disease and on brain atrophy. This issue is of particular relevance because > 90% of the subjects in this meta-analysis participated in cardiovascular trials, of which >80% were taking aspirin.

“In our opinion, it is possible that B vitamins do not delay the very slow cognitive decline observed in usual healthy aging, but the trials included in this meta-analysis cannot be used to reach this conclusion,” wrote Smith, el al.

As well, the meta-analysis authors cannot claim, as they do, that their study does not support the positive outcomes of B vitamin trials in people with mild cognitive impairment who have elevated homocysteine, given that the meta-analysis excluded trials in subjects with cognitive impairment!

To Be Meaningful, Subjects Must Be Likely to Respond

A trial or meta-analysis on the effect of B vitamins on age-related cognitive decline is only meaningful when the study population or populations are likely to respond to the intervention during the trial period — i.e., subjects who experience cognitive change detected by appropriate tests and who also have inadequate B-vitamin status.

Finally, in the Oxford University media release the authors stated the following: “Our study draws a line under the debate: B vitamins don’t reduce cognitive decline as we age. Taking folic acid and vitamin B12 is sadly not going to prevent Alzheimer’s disease.”

“Given that this meta-analysis excluded trials on cognitive impairment or Alzheimer disease, we ask the authors: ‘do they still argue that the results of this meta-analysis can be extrapolated to the effect of B vitamins on disease related cognitive impairment and dementia?’ [emphasis added]”

It is depressing that the newspaper and TV media jump on such conclusions to the detriment of health seekers everywhere. We offer thanks to American Journal of Clinical Nutrition for printing these letters. Notice, however, that there was little follow-up in the news.


  1. Clarke R, Bennett D, Parish S, et al. Effects of homocysteine lowering with B vitamins on cognitive aging: meta-analysis of 11 trials with cognitive data on 22,000 individuals. Am J Clin Nutr. 2014 Aug;100(2):657 – 66.
  2. Experts criticise ‘inaccurate’ view that B Vitamins have no role in Alzheimer’s disease prevention. St George’s University of London Web site. http://sgul.ac.uk/research/research-news/experts-criticise-inaccurate-view-that-b-vitamins-have-no-role-in-alzheimer-s-disease-prevention. Published: February 20, 2015. Accessed: February 28, 2015.
  3. Garrard P, Jacoby R. B-vitamin trials meta-analysis: less than meets the eye. Am J Clin Nutr. 2015 Feb;101(2):414 – 5.
  4. Smith AD, de Jager CA, Refsum H, Rosenberg IH. Homocysteine lowering, B vitamins, and cognitive aging. Am J Clin Nutr. 2015 Feb;101(2):415 – 6.

Featured Product

FREE Subscription

  • You're just getting started! We have published thousands of scientific health articles. Stay updated and maintain your health.

    It's free to your e-mail inbox and you can unsubscribe at any time.
    Loading Indicator