Several people have voiced concerns about the recently published Homocysteine-Lowering by B Vitamins Slows the Rate of Accelerated Brain Atrophy in Mild Cognitive Impairment: A Randomized Controlled Trial.

Citation: Smith AD, Smith SM, de Jager CA, Whitbread P, Johnston C, et al. (2010) Homocysteine-Lowering by B Vitamins Slows the Rate of Accelerated Brain Atrophy in Mild Cognitive Impairment: A Randomized Controlled Trial. PLoS ONE 5(9): e12244. doi:10.1371/journal.pone.0012244

Responding to some of the points raised by Carl Heneghan: Vitamin B and slowing the rate of brain atrophy: The numbers don't add up. I had left a series of comments but they were lost. I'll reproduce some of them here.

Heneghan:

A total of 168 participants (85 in active treatment group; 83 receiving placebo) completed the MRI section of the trial from 271. Therefore the study lost a lot of participants – where did they go?

The study authors don't really seem to have 'lost' them although reasons for declining to participate in the scans are not outlined.

A subset (187) volunteered to have cranial MRI scans at the start and finish of the study. Smith et al, 2010

It's not implausible that some of the participants might have had shunts, stents or joint replacements of such vintage that not all were readily suited to MRI investigation. Without access to the consent documents, it's not clear if some participants refused scans on the grounds that they wouldn't wish to know if they were in a group with greater brain atrophy and perhaps 'more likely' to develop AD. Unhelpfully, the appended VITACOG protocol does not clarify this matter.

We will also scan the brain by MRI before treatment and then again two years later when the treatment has finished. We can then measure the rate of brain shrinkage in each group. MRI will be voluntary for those who consent to participate in the trial and who don’t have a cardiac pacemaker.

Albeit the authors seem to have mis-judged the participation rate or there were more participants with pacemakers etc. than anticipated because the usable MRI participation rate is approx. 62%. "Participation in the MRI scans will be voluntary and we anticipate that about 80% subjects will volunteer." 187/271 indicates that 69% volunteered.

In the appended methods supplement, Smith et al state:

Eligible participants were asked if they would agree to have two cranial MRI scans, one at the start and one two years later at the end of treatment, but it was emphasized that the scans were voluntary. Although the main outcome of the trial was atrophy rate, we allowed subjects that could not or did not wish to participate in the MRI part of the study to enter the trial in a separate arm: the intention was to evaluate recruiting strategies for such trials (not reported here) for subjects with MCI.

The reference to a "separate arm" is confusing, given that only 187 of the eligible 271 appear to have consented, yet we are quoted (as per Heneghan) adverse event data from n=266.

168 of the 187 volunteers had MRIs that were suitably clear for interpretation.

I'll update this post as I try to re-construct the lost comments. However, It's not clear to me that the reported results and the confidence reposed in them are in line with the authors' description of what the study was powered to observe, as described in the protocol.

Update 1: Before addressing specific issues, I recommend. Does Industry Sponsorship Undermine the Integrity of Nutrition Research/ and Relationship Between Funding Source and Conclusion Among Nutrition-Related Science Articles

The authors provided a competing interests statement.

Dr. A. D. Smith is named as an inventor on two patents held by the University of Oxford on the use of folic acid to treat Alzheimer's disease (US6008221; US6127370); under the University's rules he could benefit financially if the patent is exploited. Drs. Refsum and A. D. Smith report having in the past received speaking honoraria from Recip AB, the company that donated the vitamin tablets, and from Axis-Shield, who make the equipment used to assay homocysteine. These competing interests do not alter the authors' adherence to all the PLoS ONE policies on sharing data and materials. None of the other authors have any financial disclosures.

It will never be clear-cut as to just how much information authors should provide in such statements. However, some Twitter exchanges have commented that Drs Smith and Refsum are associated with the contentious work of Patrick Holford. Both Smith and Refsum are members of the Scientific Advisory Board for enterprises from which Patrick Holford derives an income: eg, Food for the Brain and Brain Bio Centre.

Patrick Holford claims that Dr Smith co-taught a Food Is Better Medicine Than Drugs (FIBMTD) seminar expressly targed 'to doctors, by doctors' (see, eg., this video interview of Patrick Holford on RTE's Late, Late Show, dissected in some detail by Holfordwatch, eg. Patrick Holford and his tap-dancing on the Late Late Show). FIBMTD is a book collaboration of Patrick Holford and Jerome Burne and has its own seminar series arguing for a nutrition (but mostly in the form of supplements) approach to health problems with some notable reporting errors

Dr Smith and Patrick Holford are a regular double act in discussing research findings in newspapers such as the Daily Record before they are available for scrutiny. Dr Smith not only provided the foreward for Holford's Alzheimer's Prevention Plan book but is also said to have reviewed the manuscript (albeit, Smith's reviewing standards for non-peer review work have been criticised by HolfordWatch and Professor David Colquhoun and, to be fair, Dr Smith accepted that FFTB research for which he had been a scientific adviser and reviewer had not been  "a proper job" or "rigorous").

Update 3: It is notable that Dr Smith's informal colleague, Patrick Holford, has used the results of this paper as a marketing opportunity for homocysteine tests (not the one used in the study) and his own range of supplements (not the ones used in the study although tacitly, perhaps, endorsed by Dr Smith as part of Holford's Alzheimer's Prevention Plan (as above)). See B vitamins stop brain shrinkage preventing dementia (albeit I doubt the author meant the obvious interpretation of such a title).

In line with his own assertions, for which there is no solid evidence base, Holford advises:

The cut-off point for an effect was 10, so you want to make sure your homocysteine level is certainly below 10, although a level below 7 is optimal, especially if you are younger.

It is worth reminding readers that (the last time I looked) assays from different laboratories are not standardised and the results are not, therefore, inter-changeable or able to support the sort of interpretation or recommendation that Holford offers. As for supplementation, Holford ignores some of the concerns surrounding supplementation and the stimulation of some cancers and despite a correct summary of the study's dosages:

Smith’s team studied the effects of giving homocysteine-lowering B vitamins (folic acid 800mcg, B6 20mg and B12 500mcg)

advises:

If your level is raised, you need to supplement a homocysteine-lowering formula of B vitamins and supporting nutrients. For levels above 9, this would look something like 800mcg folic acid, 500mcg B12, 75mg B6, 20mg B2, 15mg zinc, 1.5-3g Trimethylglycine (TMG) and 500g N-Acetyl-Cysteine. If your test revels levels about 15, then you need slightly higher doses.

This last is disingenuous: Holford advises titrating vitamin supplement intake according to tHcy levels and individuals may be thus persuaded to take considerably more than Smith's team has investigated.

Update 4: Henegan has elegantly summarised one of the main causes for concern in the reporting of this study.

Although treatment with vit B after 24 months significantly slowed the rate of brain atrophy by 30% this is a relative measure. The absolute reduction is 0.32% over the length of the trial. The question is; is this clinically significance? Given adherence, was about 75% and 17/83 (21%) of the placebo group had taken supplementary folic acid or vitamin B12. In the active treatment group, and 14/ 84 (17%) in the treatment group did not take, or did not absorb, the vitamins, only 136 were defined as biologically compliant.

Until there is follow-up of this group, and a large, multi-centre trial, it is inappropriate to anticipate that the reported reductions in the rate of brain atrophy will prove to be clinically significant. Drawing upon trials in cardiovascular health, D Eva Lonn provided a useful reflection on the history of potential of the therapeutic manipulation of homocysteine levels and the, to date, disappointing clinical outcomes in populations that are already well nourished.

While the experimental and epidemiological evidence does indeed support a plausible role for homocysteine lowering in CVD prevention in the population at large (as opposed to a limited role in rare genetic disorders), overzealous interpretations of such data have led to extrapolations and unjustified early enthusiasm…

Two obvious questions arise: why did the B vitamin homocysteine-lowering trials conducted to date not demonstrate clinical benefits and is there a role for additional trials or should researchers close yet another chapter, which seemed promising but failed to deliver. The answers to these questions are complex…However, it is possible that the treatment truly has no effect on vascular risk…

In conclusion, B vitamin supplements cannot currently be recommended for the prevention of CVD events (with the exception of rare genetic disorders) and there is no role for routine screening for elevated homocysteine levels. However, ongoing clinical research should provide further evidence on whether there may be any role for homocysteine-lowering B vitamin supplements in CVD prevention and for the overall importance of homocysteine as a CVD risk factor. Lonn, E. Homocysteine-Lowering B Vitamin Therapy in Cardiovascular Prevention—Wrong Again? May, 2008; vol 299: pp. 2086-2087.

It seems that a substantial number of the participants in Smith et al.'s trial were already accustomed to taking vitamin B supplements. It might have been useful if the authors had asked the participants or the study partners to state whether they believed that they were taking the active treatment or the placebo. Depending on the formulation, B vitamins can have a distinctive taste and this might have been sufficient to 'un-blind' some of the participants in the placebo group, and may account for some of the 21% rate of additional supplementation. NB, the supplementary description of the methods mentions attention to the authentic colour of the tablets but not the taste:

The placebo tablet was identical to the vitamin tablet (TrioBe Plus®) except for omission of the vitamins and the addition of iron oxide and ferrous sulphate (0.0055%) to give a colour to match the vitamin tablet.

xx

.