Meta-analysis of the Effects of Ginkgo Extract on Behavioral and Psychological Symptoms of Dementia

Behavioral and psychological symptoms of dementia (BPSD) are characterized by symptoms of aggression, agitation, and psychosis, which typically decrease the quality of life for the patient while increasing caregiver distress, the risk for nursing home admission, and financial burden on the healthcare system. BPSD is a treatment target for patients with dementia. In conventional medicine, pharmacological management of anti-dementia drugs (i.e., acetylcholinesterase inhibitors, memantine), antidepressants, and select antipsychotics has proven beneficial for this patient population. The agitation has been observed in patients with dementia who receive the selective serotonin reuptake inhibitor (commonly referred to as “SSRI”) citalopram. EGb 761® (Dr. Willmar Schwabe GmbH & Co. KG; Karlsruhe, Germany) is a ginkgo (Ginkgo biloba, Ginkgoaceae) leaf extract. The extract contains 22.0-27.0% ginkgo flavonoids calculated as ginkgo flavone glycosides and terpene lactones, consisting of 2.8-3.4% ginkgolides A, B, and C, 2.6-3.2% bilobalide, and less than 5 ppm of ginkgolic acids. Studies show that EGb 761 is an effective treatment for BPSD. The purpose of this meta-analysis was to evaluate the effects of EGb 761 on individual BPSD; namely, an overall reduction of symptoms, reduction of symptoms present at baseline, and the prevention of newly emerging symptoms. Also, symptoms of caregiver burden were evaluated.

The following databases were searched: PubMed/Medline (from inception through December 2013), EMBASE (from January 2006 through December 2013), and PASCAL (from inception through December 2013). The search was updated June 2016. The following search terms were used: (ginkg* OR gingk*) AND clinical trial(pt) for PubMed, ((ginkg* OR gingk*) NOT medline(sb)) AND (clinical* OR trial OR randomized) for PubMed excluding Medline, (GINKGO OR GINGKO), AND (HUMAN/CT OR HOMME/CTFR) for PASCAL, and (ginkgo or gingko) AND CT = (CLINICAL TRIAL; CLINICAL STUDY; DOUBLE BLIND PROCEDURE) AND py > 2005 for EMBASE. Reference sections were screened. The manufacturer of EGb 761 was contacted for any unpublished studies. Included studies met the following criteria: (1) randomized, placebo-controlled clinical trials of EGb 761; (2) duration of ≥ 20 weeks; and (3) included patients with dementia (probable Alzheimer’s disease [AD], probable vascular dementia [VaD], or possible AD with cerebrovascular disease) and clinically significant BPSD (Neuropsychiatric Inventory [NPI] total score ≥ 6). A meta-analysis was conducted. The primary outcome measure was single-item scores on the NPI (a 12-item inventory of the presence and severity of behavioral changes in patients with dementia).

Four studies met the inclusion criteria. The authors do not say how many articles were excluded. All four studies were similarly designed (i.e., multicentered, randomized, double-blind, placebo-controlled), included patients with mild to moderate dementia, and evaluated 240 mg/day EGb 761 or placebo for 22 weeks or 24 weeks. The manufacturer of EGb 761, Dr. Willmar Schwabe GmbH & Co KG, provided individual patient data from all studies for use in the meta-analysis. The 12-item inventory designed to assess the presence and severity of altered behavior associated with dementia upon which study patients were evaluated (NPI) assessed the following signs and symptoms: delusions, hallucinations, agitation/aggression, depression/dysphoria, anxiety, elation/euphoria, apathy/indifference, disinhibition, irritability/lability, aberrant motor behavior, sleep and nighttime behavior disorders, and appetite and eating disorders. All included patients scored 35 or less on the Test for the Early Detection of Dementia with the Differentiation from Depression; had scores ranging from 9-23 on the SKT short cognitive performance test; and had clinically significant BPSD, denoted by a composite score of ≥ 6 on the NPI with at least one item score (other than delusions or hallucinations) with a value ≥ 3.

While 1628 patients were randomly assigned in the four trials evaluated, the meta-analysis included data from 1598 patients (n = 796, EGb 761; n = 802, placebo). The mean age was 66 years ± 9 years, and ≥ 67% of each group were women. There were no significant differences between groups at baseline. At the study’s end, the EGb 761 group had significantly greater improvement from baseline compared with the placebo group on the NPI composite score and caregiver distress scores (P values not reported).

  • For net effects of individual NPI composite scores, the EGb 761 group had significant improvement from baseline compared with the placebo group in the incidence of apathy (P < 0.001) > disturbance in sleep/nighttime behavior (P < 0.001) > depression (P < 0.001) > anxiety (P < 0.001) > irritability/lability (P < 0.001).

o   There were minimal changes in the incidence of delusions, hallucinations, and elation/euphoria, as well as in those endpoints that were low at baseline (i.e., disinhibition, aberrant motor behavior, and appetite/eating).

  • For net effects of individual caregiver distress scale scores, the EGb 761 group had significant improvement in the incidence of symptoms from baseline compared with the placebo group for depression (P < 0.001) > disturbance in sleep/nighttime behavior (P < 0.001) > apathy (P < 0.001) > anxiety (P < 0.001) > irritability/liability (P < 0.001).
  • Similarly, for individual NPI composite scores, the EGb 761 group had significantly greater symptom improvement from baseline compared with the placebo group for aberrant motor behavior (P < 0.001) > apathy (P < 0.001) > depression (P < 0.001) > agitation (P < 0.001) > disturbance in sleep/nighttime behavior (P < 0.001) > anxiety (P < 0.001).
  • For individual caregiver distress scale scores, the EGb 761 group had significantly greater symptom improvement from baseline compared with the placebo group for agitation (P < 0.001) > depression (P < 0.001) aberrant motor behavior (P < 0.001) > disturbance in sleep/nighttime behavior (P < 0.001) > anxiety (P < 0.001) > apathy (P < 0.001).
  • For individual NPI composite scores for symptoms not originally present at baseline but emerging during the study period, the incidence at the study’s end was significantly lower in EGb 761 group compared with the placebo group for depression (P < 0.001) > apathy (P = 0.002) > disinhibition (P = 0.001) > disturbance in sleep/nighttime behavior (P = 0.258).
  • For individual caregiver distress scale scores for symptoms not originally present at baseline but emerging during the study period, the incidence at the study’s end was significantly lower in EGb 761 group compared with the placebo group for depression (P < 0.001) > apathy (P = 0.022) > disinhibition (P = 0.004) > disturbance in sleep/nighttime behavior (P = 0.004).
  • Overall, the symptoms that were most prevalent at baseline improved in 50-60% of the EGb 761 group and 30-40% of the placebo group.

The authors conclude that 22 or 24 weeks of treatment with EGb 761 provided significant benefit for nine out of 12 symptoms of individual NPI composite and caregiver distress scores compared with placebo. The mechanism of action of how EGb 761 improves each symptom has not been confirmed but may be related to modulation of neurotransmitter systems.

This study has several strengths that make it unique and of high scientific value. (1) The authors were able to obtain raw data from all studies to conduct the meta-analysis. (2) The four included studies were homogeneous. (3) The population size was relatively large. (4) All included studies evaluated the same dose and preparation of ginkgo.

This article also had some limitations. (1) The authors did not report how many articles were located in their original search. (2) The authors did not provide P values of the total NPI composite and caregiver distress scores. (3) The search was English-language only so some studies may have been missed. (4) Considering that the authors had access to raw data, it would have been beneficial if they had conducted a meta-analysis of safety. An analysis of risk/benefit is always of value and this study missed that opportunity.

Two of the authors (Mueller and Hoerr) are employees of Dr. Willmar Schwabe GmbH & Co. KG, and the others have received speakers’ honoraria.


Savaskan E, Mueller H, Hoerr R, von Gunten A, Gauthier S. Treatment effects of Ginkgo biloba extract EGb 761® on the spectrum of behavioral and psychological symptoms of dementia: a meta-analysis of randomized controlled trials. Int Psychogeriatr. September 21, 2017:1-9. doi: 10.1017/S1041610217001892.



Meta-analysis Finds Standardized St. John’s Wort Extracts as Effective as Conventional Antidepressants

  • St. John’s Wort (Hypericum perforatum, Hypericaceae)

St. John’s wort (SJW; Hypericum perforatum, Hypericaceae) aerial parts are a popular treatment for depression, and many countries in Europe prescribe SJW for that purpose. According to the authors, SJW has been well researched; however, the results are conflicting. The last large published meta-analysis was conducted in 2008, and it found SJW superior to placebo and similar to conventional antidepressants. Since then, additional research has emerged. Hence, the objective of this systematic review and meta-analysis was to provide an updated analysis.

The following databases were searched from January 1, 1960, through May 1, 2016, according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines: PubMed; Ovid; Clinical Trials Register of the Cochrane Collaboration Depression, Anxiety and Neurosis Group (CCDANTR); Cochrane Field for Complementary Medicine; China National Knowledge Infrastructure; and Wanfang. The following search terms were used: [St John’s Wort OR Hypericum perforatum OR hypericin OR hyperforin OR johannes kraut (German for St John’s wort) OR 圣约翰草 (Chinese for St John’s wort)] AND [depression OR antidepressant OR SSRI]. Reference lists of articles also were searched.

Studies were included if the following criteria were met: (1) were randomized controlled trials; (2) compared a standardized extract of SJW with a selective serotonin reuptake inhibitor (SSRI); (3) included patients clinically diagnosed with major depressive disorder according to criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV), International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10), or Chinese Classification of Mental Disorders, Third Edition (CCMD-3); (4) had extractable outcome measures for treatment efficacy and safety; and (5) included ≥ 20 human subjects. The primary outcome measures were treatment response and safety/incidence of adverse effects (this was evaluated as dropout rate). Treatment response was assessed with (1) the response rate (or relative risk [RR]) ratio of the proportion of responders in the treatment group versus the control group, with response defined as reduction in Hamilton Rating Scale for Depression (HAM-D) score > 50% and (2) remission rate ratio, which is the proportion of remitters in the treatment group versus the control group, with remission defined as reduction in HAM-D score > 75%. The Jadad scale and the Cochrane Collaboration’s tool for assessing the risk of bias were used to assess methodology. Heterogeneity was tested statistically, and sensitivity analyses were conducted.

A total of 5428 articles were located and 27 articles met the inclusion criteria. Included studies were from China, Germany, Denmark, Brazil, and the United States. Studies from China had weaker methodological quality and a higher risk of bias, such as inadequate generation of a randomized sequence (selection bias), lack of blinding of outcome assessment (detection bias), and incomplete outcome data (attrition bias). Based on statistical analysis, the likelihood of publication bias was small. The studies included 40 to 428 patients with mild-to-moderate depression. The studies evaluated the following SJW extracts: WS® 5570 (Dr. Willmar Schwabe GmbH & Co. KG; Karlsruhe, Germany), WS® 5572 (Dr. Willmar Schwabe GmbH & Co. KG), Ze 117 (Zeller AG; Romanshorn, Switzerland), STW3 (Steigerwald Arzneimittelwerk GmbH; Darmstadt, Germany), STW3-VI (Steigerwald Arzneimittelwerk GmbH), LoHyp-57 (Dr Werner Loges and Co., GmbH; Winsen, Germany), and LI-160 (Lichtwer Pharma; Berlin, Germany). They contained standardized 0.3% hypericin and 2-5% hyperforin, with doses of 300-1350 mg/day. The duration of treatment ranged from four to 12 weeks.

According to the meta-analysis, SJW had a similar response rate to SSRIs (pooled RR, 0.983; 95% confidence intervals [CI], 0.924-1.042; P < 0.001). Similar results were found when using a subgroup analysis of only trials with good methodological scores (Jadad scale of 3 or higher). SJW also had similar remission rates to SSRIs (pooled RR, 1.013; 95% CI, 0.892-1.134; P < 0.001). SJW had a significantly lower discontinuation/dropout rate compared to SSRI treatment (pooled odds ratio, 0.587; 95% CI, 0.478-0.697; P < 0.001).

The authors conclude that the “findings further strengthen the support for St John’s wort’s clinical efficacy in reducing depressive symptoms and more conclusively showed that St John’s wort had a significantly smaller number of patients discontinuing treatment/dropping out due to adverse/side effects. This is significant as discontinuation of antidepressants by patients is a common problem faced in the clinical setting, and greatly hampers the success of treatment.” The conclusions of this study support the 2008 Cochrane review1; however, the present meta-analysis includes more recent studies, larger studies, and studies published in the Chinese literature, thereby extending the findings of the Cochrane review. The authors acknowledge a lack of consensus regarding the appropriate dosage of SJW; however, they did not evaluate the doses used in this meta-analysis. It would have been beneficial if they could have done a sub-analysis to elucidate the optimal dose. Based on the availability of the research, the authors acknowledge that longer-term studies and studies in children < 18 years are needed. This study was very well designed, conducted, and written. It is unique in that it did not exclude non-English articles. The authors report no conflict of interest.


1Linde K, Berner MM, Kriston L. St John’s wort for major depression. Cochrane Database Syst Rev. October 8, 2008;(4): CD000448. doi: 10.1002/14651858.CD000448.pub3

Ng QX, Venkatanarayanan N, Ho CYX. Clinical use of Hypericum perforatum (St John’s wort) in depression: A meta-analysis. J Affect Disord. 2017;210:211-221.