Skip navigation

Ginkgo for dementia

Clinical bottom line: No convincing beneficial effect of relatively low doses of ginkgo bilboa extract on dementia was shown. Higher doses were effective in patients with mild to moderate disease, but the analyses were based on few patients. Studies were mainly of short duration, assessed different doses of gingko and different durations of treatment.

Systematic review:

Ernst E, Pittler MH. Ginkgo bilboa for dementia. A systematic review of double-blind, placebo-controlled trials. Clin Drug Invest 1999: 17: 301-308.

Date review completed: 1998

Number of trials included: 9

Number of patients: (891 total)

Control groups: placebo.

Main outcomes: Clinical global impression, various geriatric scales, Alzheimer's disease assessment scale.

Inclusion criteria were randomised, double blind, placebo controlled trials which assessed the effects of ginkgo bilboa on dementia (Alzheimer, multi-infarct or mixed type dementia). Studies on cerebral insufficiency were excluded.

Medline, Embase, Biosis and the Cochrane Library were searched (to March 1998) for published reports. Bibliographies of retrieved reports were checked for additional citations and no language restrictions were made. Manufacturers of gingko preparations were contacted for published and unpublished information. Data were extracted in a standardised, predefined manner by the two reviewers. A descriptive analysis was conducted.



The included studies were of good methodological quality (minimum score of 3 on a validated 5-point quality rating scale). All assessed standardised ginkgo bilboa extract in patients with mainly mild or moderate dementia. Most studies were of 4-12 weeks duration; two large trials were longer (24 weeks and one year). The reviewers' did not mention the number of patients improved in the individual trials.

Duration - over six months:

One study (327 patients) assessed oral ginkgo 120 mg daily in patients with mild to moderately severe Alzheimer's Disease or multi-infarct dementia over one year. Only 137 patients completed the trial (50% for ginkgo, 38% for placebo). Significant improvements with ginkgo compared with placebo were shown for ratings with the Alzheimer's Disease Assessment Scale and a Geriatric Evaluation (per protocol). No difference was shown for the clinical global impression.

The other study assessed oral ginkgo 240 mg daily in 222 patients aged 55 years or older with mild to moderate Alzheimer's Disease or multi-infarct dementia over 24 weeks. All outcomes, including the clinical global impression showed a significantly greater effect with ginkgo than with placebo. The proportion of patients who responded to ginkgo was 28% and with placebo it was 10%.

Duration - up to 3 months:

Six studies (302 patients) assessed oral ginkgo 120-240 mg daily compared with placebo in patients with a variety of dementia disease states over 6-12 weeks. Three studies showed significantly greater improvement with gingko than with placebo and three showed no significant benefit.

One trial assessed intravenous gingko 200 mg four times per week for four weeks in 40 patients with mild or moderate Alzheimer's Disease, multi-infarct, or mixed type dementia. Significantly greater improvement was shown with intravenous ginkgo than with placebo for all outcomes assessed. For the clinical global impression, improvement was reported in 85% of patients treated with gingko and 35% treated with placebo.

Adverse effects

Adverse effects were uncommon. Gastrointestinal complaints and headache were the most commonly reported adverse effects. Allergic skin reactions were reported in one trial with oral ginkgo 240 mg.



There is no convincing evidence to show that lower doses of ginkgo bilboa extract improve dementia. However, oral ginkgo 240 mg and intravenous ginkgo 200 mg did produce significantly greater improvements than placebo. The superiority of the intravenous dose was based on only 40 patients, so the result is not robust. One problem in these trials is their ability to demonstrate a difference between study treatments, since patients with mild dementia were included. If patients with only moderate or severe dementia had been included the internal sensitivity of the studies would have been greater. This is an important measure of study validity. Another problem in these trials is that different outcomes, doses and durations of treatment were assessed. The studies were clinically heterogeneous and patient information could not be pooled in a meta-analysis.

Further reading

Related topics


Dementia - diagnosis and treatment


AT123 - 4648 5914 Gingko for dementia: Aug-2000