July 22, 2011

expert reaction to study of B-vitamin treatment and cognitive decline

A small clinical trial showed that B vitamins may be linked to the slowing of cognitive decline in people with mild cognitive impairment.

Robert Clarke, Senior Research Fellow, at the Clinical Trial Service Unit of the University of Oxford, said:

“The VITACOG trial of 266 older people did not demonstrate any overall effect of B-vitamins on cognitive function, and the reported effects were only identified in a post-hoc sub-group analysis of people with high homocysteine levels. In small trials, ‘significant results’ are often generated by such sub-group analyses, and may arise purely by chance. As such, the implied effect of B-vitamins may well be spurious. By contrast, the SEARCH trial reported in 2010 that B-vitamins had no significant on cognitive function in over 7000 UK patients treated for up to 7 years (JAMA 2010, 303: 2486-94).”

Paul Matthews, Professor of Clinical Neurosciences, Imperial College, London, said:

“Previous trials of vitamin B12 and folate have largely been negative, but they also have been relatively small, so a modest treatment effect could have been missed. However, the current study was not designed to test any of the measures presented – these are secondary measures to the original study. The results therefore should only be used to suggest new questions for a further study. The CLOX outcome also is not a typical primary measure for Alzheimer’s disease – while impairments can be found, this is not a well-validated measure and certainly is not specific for AD. Furthermore, the subgroup analyses are interesting but impossible to assess with confidence. If the authors’ primary hypothesis were correct, it is not clear why the CLOX subgroup analysis was not positive. The subgroups are very small and there are so many ways of composing sub-groups, it is hard to escape the bias that a sub-grouping that ‘works’ is highlighted, while one that does not is ignored. The small sizes mean that, like the study overall, they are subject to biases because the subjects may not be matched for all key factors. We have no information on this.

“Overall, this interesting report of secondary measures strengthens a rationale for further investigations. Some people who complain of memory problems and who have vitamin B12 or folate levels in their blood within what is usually considered a normal range for the population may benefit from supplementation. As a community, we are set important challenges: Who are the people who might realise any benefit? Will benefits suggested by the specialised tests translate into meaningfully better maintenance of good memory and attention?”

John Hardy, Professor of Neuroscience at University College London , said:

“These are interesting results, but as the authors acknowledge, this is a small trial and one has to be concerned that the results will not generalise. Clearly, though, the results suggest a follow up larger trial would be valuable.”

Dorothy Bishop, Professor of Developmental Neuropsychology, University of Oxford, said:

“Overall, I find these results suggestive but not particularly compelling. The most interesting reported effects result from post-hoc subgroup analysis. My usual reaction to significant effects that only emerge on this basis is that they are probably chance fluctuations but if they are novel and consistent with theory, they may indicate that it is worth doing further research on the topic. The problem here is that the further research that has already been done is not encouraging. This small study should not be given undue prominence in comparison to existing data with less marketable implications.”

Dr Carl Heneghan, Director of the Centre for Evidence-Based Medicine, University of Oxford, said:

“This was a small, double-blind, single-centre study, which included participants with mild cognitive impairment (MCI).

“The efficacy analyses were performed on the ‘intention-to-treat’ population, defined as the population who were randomly assigned to treatment and who received at least one dose of study medication. In the whole intention-to-treat cohort, there was no significant effect of B vitamins on CDR (P=0.23) or IQCODE (P=0.26).

“The authors state in the discussion that the vitamin didn’t work: ‘In the intention-to-treat cohort of participants with MCI who completed the trial, B-vitamin treatment did not improve performance in tests of global cognitive function (MMSE), episodic memory (HVLT-DR) or semantic memory (category fluency) or on measures of clinical status (CDR and IQCODE).’ Elsewhere they state ‘However, B-vitamin treatment did stabilise performance on the CLOX test of executive planning function.’ This is likely to be a chance result and is one subgroup of many.

“Results based on subgroups are problematic. I draw your attention to the article by Peter Rothwell in the Lancet on subgroup analysis in randomised control trials: (http://www.ncbi.nlm.nih.gov/pubmed/15639301).
‘Analyses must be predefined, carefully justified, and limited to a few clinically important questions, and post-hoc observations should be treated with scepticism irrespective of their statistical significance.
‘If important subgroup effects are anticipated, trials should either be powered to detect them reliably or pooled analyses of several trials should be undertaken.’

“In this study the subgroups are too small and there are too many to make the results meaningful, which the authors note in the limitations section: ‘A limitation of the trial is the small sample size’.”

‘Cognitive and clinical outcomes of homocysteine-lowering B-vitamin treatment in mild cognitive impairment: a randomized controlled trial’ by Celeste de Jager et al., published in the International Journal of Geriatric Psychiatry on Friday 22 July 2011.