Sep 17 2012

No Benefit from Ginkgo biloba in MS

One of the themes of this blog (and my other medical blog, science-based medicine) is that there is a structure and natural history to scientific (and specifically medical) research and in order to understand the answer to any specific scientific question one must look at the whole of the research, not just a single study.

Analyzing individual studies is important because they are the units of which the scientific literature is comprised. Further, some individual studies are large, rigorous, and fairly definitive – but it takes a long time to get there, and most of the scientific literature is comprised of less-than-definitive studies.

There are also recurring patterns in the research that help us put individual studies into context and better arrive at reliable conclusions, which is the whole point of research in the first place. For example, medical studies usually begin with pre-clinical basic science, then progress to pilot clinical studies. A pilot study is small and usually less rigorous in design. Such studies are exploratory – their purpose is to see if we should even bother, and if it will be safe, to do larger more difficult trials. Studies progress with larger or better designed studies until we get to fairly definitive trials. Then and only then do we have some idea if a treatment actually works and is safe.

It can often take 1-2 decades, however, to get to that point. Meanwhile a vast body of preliminary exploratory research may have been generated, with systematic reviews and meta-analysis to help us make sense of it all. The preliminary exploratory stage of scientific research is often little more than a Rorschach test – people see in it what they want.

In medical research there is a well-documented positive bias to preliminary research. There is researcher bias combined with researcher degrees of freedom (methods of subconsciously biasing a study to the desired outcome), placebo effects, and publication bias, all conspiring to make a treatment seem as if it works when it doesn’t.

The story of Ginkgo biloba fits into this general pattern – early weakly positive studies showing a potential benefit from Ginkgo for memory symptoms, and even dementia and Alzheimer’s disease. A 2009 large rigorous study, however, found no benefit at all. Since Ginkgo is a popular herbal remedy its use was based more on tradition (which sometimes means someone’s invented sales hype) rather than solid basic science. There never was much of a rationale for thinking Ginkgo would help in memory, beyond common use. In any case, the clinical evidence is ultimately what mattered, and it showed that Gingko has no benefit.

The same pattern has now been repeated with regard to Ginkgo and specifically the treatment of cognitive symptoms with multiple sclerosis (MS). A pilot study involving only 22 individuals (a good rule of thumb is that any study with less than 50 subjects should be considered preliminary) showed a possible benefit from Gingko in MS, specifically on reports of fatigue and functional performance. The study concluded:

This exploratory pilot study showed that no adverse events or side effects were reported and that ginkgo exerted modest beneficial effects on select functional measures (eg, fatigue) among some individuals with MS.

A second study involving 38 subjects found no statistically significant results, but did find a non-significant trend in improvement in the Stroop test. The authors concluded that the data: “suggests that GB may have an effect on cognitive domains assessed by this test.”

The whole point of pilot studies, as I indicated above, is not as a basis for clinical treatment but as a guide to further research. The study provided some evidence that Gingko is at least safe and may have a benefit, so researcher performed a follow up study, which has now been published. This study involved 120 subjects divided into two groups, one receiving Ginkgo (120mg twice a day) and the other placebo. They used four measures of cognitive ability at baseline and after 12 weeks and found no statistically significant difference in any measure.

It’s hard to make a convincing trend from three studies, but they do mirror the pattern seen with Ginkgo studies in general – the larger the study the smaller the effect and the most rigorous studies show no effect. We also see the general trend of a positive bias among preliminary studies.

Ginkgo is still widely sold with claims that it improves cognition, despite the negative evidence. It’s hard to find exact figures without paying for market research, but I find estimates are all in the range of hundreds of millions of dollars annually in sales worldwide.

The story of Ginkgo is typical – early positive-biased research with later more rigorous studies finally giving us reliable information, in this case that Gingko is ineffective for the indication for which it is most commonly used.

5 responses so far