Wouldn’t it be nice if people could pop a pill to prevent dementia? If ever it comes to pass, that pill won’t contain B vitamins, judging from the results of research that appeared in the November 12 Neurology. Scientists led by Lisette de Groot, Wageningen University, the Netherlands, found that a supplement containing vitamin B12 and folic acid (vitamin B9) did not improve cognitive performance in nearly 3,000 volunteers predicted to get a boost from the supplements. These older adults had high blood levels of a neurotoxic amino acid called homocysteine, which the vitamins help to metabolize. “It does not rule out [a benefit] in people who are cognitively impaired or who have low vitamin B levels,” said corresponding author Rosalie Dhonukshe-Rutten.

Homocysteine is a sulfur-containing amino acid known to damage neurons. People with high plasma homocysteine or low vitamin B levels reportedly carry a greater risk of dementia (see Seshadri et al., 2002, and Agnew-Blais et al., 2014). However, randomized controlled trials demonstrate little cognitive benefit from reducing homocysteine levels with vitamin B supplements in healthy older adults (see Clarke et al., 2014Ford and Almeida, 2012). Since many of these previous studies ran for less than a year, were small, and rarely selected for high homocysteine levels, the authors aimed to reassess the supplements over two years in a large population who had high homocysteine levels.

First author Nikita van der Zwaluw and colleagues enrolled 2,919 Dutch men and women over the age of 65 in the B-Vitamins for the Prevention of Osteoporotic Fractures study (B-PROOF). The main goal was to see if the supplements prevented bone fractures, but since vitamin B has been associated with various other benefits, the authors tacked on additional assessments—among them, cognitive tests. About half of the participants received a placebo pill, while the other half took a daily supplement of 400 μg folic acid and 500 μg vitamin B12 over two years.

The authors selected participants with plasma homocysteine levels of 12-50 μmol/L. Less than 10 μmol/L is considered normal. At baseline and after treatment, the researchers used the Mini-Mental State Examination (MMSE) and Rey Auditory Verbal Learning Test to measure global cognitive performance and episodic memory, respectively. A subset of a little more than 700 patients also completed tests of attention and working memory, information-processing speed, and executive function. All gave blood samples before and after the study so that the researchers could measure homocysteine.

Over the treatment period, plasma homocysteine levels dropped 1.3 μmol/L in the placebo group and 5 μmol/L in the treatment group. The larger decline in those taking B vitamins did not translate to a cognitive benefit. They scored slightly better on the MMSE, but the difference was insignificant after adjusting for practice effects. In post hoc analyses, the participants in the treatment group who had the lowest initial levels of plasma vitamin B12 slightly improved their processing speed.

The results fit in with a slew of other cognitive data regarding vitamin B supplements in healthy elderly people. Most clinical trials suggest that supplements lower homocysteine levels without improving cognitive health, said Murray Skeaff, University of Otago, New Zealand, who co-authored a recent meta-analysis of 11 such trials (see Clarke et al., 2014). Paul Aisen, University of California, San Diego, agreed. “Homocysteine reduction [once] seemed to be an attractive therapeutic goal, but this study provides further evidence that this strategy fails to produce beneficial effects on cognitive performance,” he wrote to Alzforum in an email. “Our efforts should be directed elsewhere.”

However, the authors insist that even two years may have been too short to see a cognitive benefit. David Smith, University of Oxford, U.K., noted that on the whole, these participants started out with good nutrition and the placebo group declined little on cognitive tests. “One would not expect much, if any, effect of giving B vitamins to this cohort,” he wrote to Alzforum in an email (see full comment below).

Sudha Seshadri, Boston University, agreed that the findings were not definitive. “The study didn’t have an opportunity to see an effect,” she told Alzforum. Both Seshadri and Smith echoed the authors’ notion that these results have no bearing on people who are vitamin-deficient or cognitively impaired. Future studies should focus on older adults with some cognitive impairment, or vitamin deficiency, or should use more sensitive cognitive tests to detect changes in healthy people, Seshadri said. “I don’t think this is the last word on the subject,” she said.

While the authors will not follow these participants further, they plan to analyze magnetic resonance imaging data from a subset of about 200 of them, taken at the two-year mark, to see if the supplements associate with higher average brain volume.—Gwyneth Dickey Zakaib


  1. In this paper by van der Zwaluw and colleagues, there are some points to consider. First, in eight out of 12 cognitive tests there was no cognitive decline in the placebo group over two years: There was either no change, or even an improvement, and so it is difficult to know what effect of B vitamins was expected. In four of the tests there was decline in the placebo group and in one test (MMSE) this was significantly reduced by B vitamin treatment. But the changes in the other tests were very small and the authors did not provide a power estimate of whether there were enough subjects to detect a significant effect of B vitamin treatment. It should be noted that the whole cohort was only used for MMSE and episodic memory tests. The other tests were done in a subgroup of between 721 and 759 subjects.

    Given these caveats, the authors cannot claim that they have “provided Class I evidence that two-year supplementation with B vitamins … does not affect cognitive performance,” because there was either no cognitive decline in the placebo group, or when there was decline it was small and it is likely that the number of subjects and/or the duration of the trial would not have been sufficient to detect a treatment effect. In other words, the trial was probably underpowered. The authors themselves note other limitations of the study, in particular the lack of an intermediate testing session, which prevents any modelling of longitudinal change and also means that the fairly large number of drop-outs (422) cannot be included in the analysis.

    The authors reported that in a post-hoc analysis the subjects with low baseline holotranscobalamin, a measure of vitamin B12 sufficiency, showed a beneficial effect of B vitamin treatment on episodic memory. This result is consistent with the view that this was on the whole a very healthy population mostly with a good B vitamin status and so one would not expect much effect of extra B vitamins, except in the subgroup with poorer status.

    It would have been interesting to look at the subgroup in the cohort who did show cognitive decline and test by logistic regression to see if the relative risk of cognitive decline was influenced by B vitamin treatment.

    The authors correctly pointed out in their last paragraph that “we cannot extrapolate this [result] to persons who already have cognitive problems.” So readers should not assume that the results are relevant to the possible treatment of cognitive impairment.

    Previous studies on normal elderly, many of which were included in Clarke’s recent meta-analysis, were consistent about one thing: In the placebo groups there was little, if any, cognitive decline. This was mainly for two reasons: the groups studied were very healthy, with good nutrition and so good B vitamin status on the whole; secondly, the tests used (such as MMSE) were far too insensitive for longitudinal studies. For example, MMSE declines about 0.1 point per year so you need a very long trial to show an effect. The published trials were all too short and/or underpowered in terms of number of participants. However, in those subjects who either start with poor vitamin status or who are already cognitively impaired (as in mild cognitive impairment, MCI) the story is different. In Durga’s FACIT trial (see Durga et al., 2007) the subjects recruited on the basis of having high homocysteine and then folic acid did indeed slow cognitive decline. In the WAFACS trial (Kang et al., 2008), women with low B vitamin status at the start showed slowing of decline by B vitamin treatment. Even the Aisen trial (see Aisen et al., 2008) in people with AD was not as negative as people think: In a subgroup analysis they reported that those with mild AD did show slowing of cognitive decline over 15 months with B vitamins, whereas those with moderate AD did not show any effect.

    So that could be indicative of a critical window: Those in whom the disease has progressed too far will not benefit, while those who have not started to decline cannot benefit because there is no decline to slow down. Only those in between, typified by people with MCI or mild AD, will show benefit. In these people, the high homocysteine is driving more rapid brain atrophy which, in turn, is driving cognitive decline. We recently showed this causal pathway by Bayesian network modelling (Douaud et al., 2013). 

    Whether or not normal people with high homocysteine who do not show any cognitive impairment will benefit from lowering homocysteine is unanswered: For that we need larger and much longer clinical trials. In the Oxford Project to Investigate Memory and Ageing (OPTIMA) cohort we observed that high homocysteine in controls increased the risk of progressing to MCI, but that took at least eight years (A. Oulhaj et al., unpublished). The famous Gothenburg women’s study by Skoog and colleagues (see Zylberstein et al., 2011) showed that raised homocysteine was a strong risk factor for dementia 35 years later. More trials are needed because, if it is really true, then a lot of people will benefit from increasing B vitamin intake, especially B12. The population-attributable risk of AD due to raised homocysteine is 22 percent and the prevalence of raised homocysteine is about 30 percent (see Beydoun et al., 2014). Fundamentally, there is a difference between the slow cognitive decline of normal aging and the more rapid decline due to a disease process: Homocysteine is one factor in the latter but we do not yet know if it is involved in the former.


    . Effect of 3-year folic acid supplementation on cognitive function in older adults in the FACIT trial: a randomised, double blind, controlled trial. Lancet. 2007 Jan 20;369(9557):208-16. PubMed.

    . A trial of B vitamins and cognitive function among women at high risk of cardiovascular disease. Am J Clin Nutr. 2008 Dec;88(6):1602-10. PubMed.

    . High-dose B vitamin supplementation and cognitive decline in Alzheimer disease: a randomized controlled trial. JAMA. 2008 Oct 15;300(15):1774-83. PubMed.

    . Preventing Alzheimer's disease-related gray matter atrophy by B-vitamin treatment. Proc Natl Acad Sci U S A. 2013 Jun 4;110(23):9523-8. PubMed.

    . Midlife homocysteine and late-life dementia in women. A prospective population study. Neurobiol Aging. 2011 Mar 1;32(3):380-6. PubMed.

    . Epidemiologic studies of modifiable factors associated with cognition and dementia: systematic review and meta-analysis. BMC Public Health. 2014 Jun 24;14:643. PubMed.

    View all comments by A. David Smith

Make a Comment

To make a comment you must login or register.


Paper Citations

  1. . Plasma homocysteine as a risk factor for dementia and Alzheimer's disease. N Engl J Med. 2002 Feb 14;346(7):476-83. PubMed.
  2. . Folate, Vitamin B-6, and Vitamin B-12 Intake and Mild Cognitive Impairment and Probable Dementia in the Women's Health Initiative Memory Study. J Acad Nutr Diet. 2015 Feb;115(2):231-41. Epub 2014 Sep 8 PubMed.
  3. . 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 Jun 25;100(2):657-666. PubMed.
  4. . Effect of homocysteine lowering treatment on cognitive function: a systematic review and meta-analysis of randomized controlled trials. J Alzheimers Dis. 2012;29(1):133-49. PubMed.
  5. . Rationale and design of the B-PROOF study, a randomized controlled trial on the effect of supplemental intake of vitamin B12 and folic acid on fracture incidence. BMC Geriatr. 2011 Dec 2;11:80. PubMed.

Further Reading


  1. . Associations between medication use and homocysteine levels in an older population, and potential mediation by vitamin B12 and folate: data from the B-PROOF Study. Drugs Aging. 2014 Aug;31(8):611-21. PubMed.
  2. . The association between plasma homocysteine levels and bone quality and bone mineral density parameters in older persons. Bone. 2014 Jun;63:141-6. Epub 2014 Mar 12 PubMed.
  3. . Folate, Vitamin B-6, and Vitamin B-12 Intake and Mild Cognitive Impairment and Probable Dementia in the Women's Health Initiative Memory Study. J Acad Nutr Diet. 2015 Feb;115(2):231-41. Epub 2014 Sep 8 PubMed.
  4. . The worldwide challenge of the dementias: a role for B vitamins and homocysteine?. Food Nutr Bull. 2008 Jun;29(2 Suppl):S143-72. PubMed.
  5. . B vitamin supplementation improves cognitive function in the middle aged and elderly with hyperhomocysteinemia. Nutr Neurosci. 2014 Jun 18; PubMed.

Primary Papers

  1. . Results of 2-year vitamin B treatment on cognitive performance: secondary data from an RCT. Neurology. 2014 Dec 2;83(23):2158-66. Epub 2014 Nov 12 PubMed.