An independent systematic review of the world’s medical literature has concluded that there is little evidence to link the popular dietary supplement made with the standardized extract of ginkgo leaf with cases of bleeding.
After reviewing all of the available literature on ginkgo, the authors concluded that the evidence for ascribing causality of bleeding to ginkgo use was “uncertain.” The researchers also concluded that since there have been no cases of bleeding observed in 9772 subjects in 44 controlled clinical trials that they reviewed, that the probability of such cases happening with ginkgo users was “unlikely.”
Medical experts at the University of Exeter in England searched 5 medical databases for cases reports in which bleeding had been observed in patients who had been using some type of preparation containing the popular herb ginkgo. They found 41 articles up to October 2003 (when the initial searches were conducted) of which 12 cases reports met the authors’ inclusion criteria. They examined all of the details available from each of these published reports and scored the reports according to a previously published reliability scoring scale, based on the amount of evidence in each report. The resulting scores were: “unevaluable --the report contains inadequate information to assess the likelihood of a causal relationship”; “possible - the report provides some evidence for a causal relationship but there may be other causes of the event” (e.g., the use of anti-coagulant drugs); and “likely - the report is well documented and appears to provide reliable evidence for a causal relationship.”
In only 2 of the cases was the nature of the ginkgo extract adequately identified. This has been a problem for many years in such case reports; that is, medical professionals have not included brand names and clear descriptions of herbal products in case reports in which the herbal supplements have been implicated in an adverse event.
The authors concluded that the evidence in the reports was “far from compelling.” They noted that in all but one of the cases the evidence for causality did not rate higher than “possible”. They noted that with such widespread use of ginkgo - 5 million unit doses are sold each year in Germany alone - the number of case reports of bleeding was “extremely low.”
The authors discuss the apparent contradiction in the case reports and the results from controlled clinical trials. “Case reports can rarely be conclusive,” they write, “and under-reporting can significantly distort the picture….While the case reports suggest that [ginkgo] does affect blood coagulation, the controlled clinical trials do not support such a hypothesis. Weighing the conclusiveness of this evidence, it seems likely that the case reports are ‘false positives’ and that the controlled trials depict the true situation: G. biloba does not cause bleeding abnormalities.”
The review was initiated in recognition that numerous published articles and reviews on ginkgo have acknowledged the theoretical possibility that use of ginkgo extract as a medicine or dietary supplement may result in bleeding due to ginkgo’s well-documented mechanism of inhibiting platelet aggregation. Platelets are structures in the bloodstream that clump together as a natural defense mechanism to help stop bleeding. Some compounds, like the chemicals in ginkgo, are known to inhibit this clumping or aggregation process, and thus, the concern that ginkgo may interfere with the natural aggregation process or possibly even cause bleeding.
According to Professor Edzard Ernst of the University of Exeter, one of the studies authors, and a prolific author of meta-analyses and systematic reviews of the research on herbs, phytomedicines, and complementary and alternative medicine modalities, “All books on the subject warn that ginkgo causes bleeding - based on the evidence, I think they are wrong.” [Ernst E. Personal communication to M. Blumenthal, Mar. 14, 2005.]
Ernst E, Canter PH, Coon JT. Does Ginkgo biloba increase the risk of bleeding? A systematic review of case reports. Perfusion 2005;18-52-56.