Papers

Effectiveness of Ginkgo biloba in treating tinnitus: double blind, placebo controlled trial

Shelley Drew, research associate, Ewart Davies, senior lecturer. 

Pharmacology Department, Division of Neuroscience, University of Birmingham, Birmingham B15 2TT

Correspondence to: S Drew s.j.drew{at}bham.ac.uk

	Abstract

Top Abstract Introduction Participants and methods Results Discussion References

Objective: To determine whether Ginkgo biloba is effective in treating tinnitus.
Design: Double blind, placebo controlled trial using postal questionnaires.
Participants: 1121 healthy people aged between 18 and 70 years with tinnitus that was comparatively stable; 978 participants were matched (489 pairs).
Intervention: 12 weeks' treatment with either 50 mg Ginkgo biloba extract LI 1370 three times daily or placebo.
Main outcome measures: Participants' assessment of tinnitus before, during, and after treatment. Questionnaires included items assessing perception of how loud and how troublesome tinnitus was. Changes in loudness were rated on a six point scale. Changes in how troublesome were rated on a five point scale.
Results: There were no significant differences in primary or secondary outcome measures between the groups. 34 of 360 participants receiving active treatment reported that their tinnitus was less troublesome after 12 weeks of treatment compared with 35 of 360 participants who took placebo.
Conclusions: 50 mg Ginkgo biloba extract LI 1370 given 3 times daily for 12 weeks is no more effective than placebo in treating tinnitus.

	Introduction

Top Abstract Introduction Participants and methods Results Discussion References

Tinnitus, or "ringing in the ears," is a common condition recognised as a problem by about 10% of the population and considered a major problem by about 0.5%.¹ There are no effective pharmacological treatments for tinnitus. Because tinnitus is considered to have a number of underlying causes, it is unlikely that a single treatment will be effective for all patients. Therefore, trials of treatments for tinnitus need to be capable of identifying treatments that may help only a subgroup of those with tinnitus. Such trials should be well controlled and include large numbers of patients. Previous trials have failed to meet these criteria and have produced inconsistent and ambiguous results.²

Extracts from the Ginkgo biloba tree have been used in Chinese medicine for thousands of years. Recently, however, Ginkgo biloba extracts have become commonly available in health food stores throughout the United Kingdom; Ginkgo biloba is one of the top 10 selling herbs in health food stores in the United States.³ High quality, standardised extracts from the leaves of the tree have been shown to have significant therapeutic effect on the symptoms of cerebral insufficiency, including memory disturbances and other cognitive deficits such as tinnitus. ^{4 5}

Prospective studies carried out to determine whether the extract is effective in treating tinnitus without accompanying symptoms of cerebral insufficiency have provided inconsistent results.² None the less, Ginkgo biloba is frequently suggested as a possible treatment for tinnitus in the press, and many people with tinnitus are using a variety of products on the basis of limited evidence.

In this study a standardised extract of Ginkgo biloba (LI 1370, Lichtwer Pharma, Berlin, Germany) was used in a large, controlled trial to determine whether it is effective in treating tinnitus. This is one of the most popular brands sold in the United Kingdom, and the extract conforms to the requirements of the German Commission E monograph.⁶

	Participants and methods

Top Abstract Introduction Participants and methods Results Discussion References

Participants
Participants were recruited through advertisements in the national press in the United Kingdom and the British Tinnitus Association's publication, Quiet. Altogether, 1121 participants were selected from the original 8667 applicants and matched when possible for sex, age (10 years difference), and the duration of tinnitus (5 years). (Details of the exclusion criteria can be found on the BMJ 's website.)

Methods
This double blind, placebo controlled trial was carried out entirely by mail and telephone. All procedures were approved by the local ethics committee (South Birmingham Health Authority). Assuming that there would be a significant improvement in tinnitus in 30% of participants taking placebo, the calculations predicted that it would be necessary to have 496 patients in each group to show a 10% improvement over placebo among those taking active treatment with a power of 90% at the 0.05 significance level.⁷ The sample size was set to account for withdrawals.

Intervention
The treatment was provided as 252 tablets containing 50 mg of either Ginkgo biloba standardised extract LI 1370 (containing 25% flavonoids, 3% ginkgolides, and 5% bilobalides) or placebo (both provided by Lichtwer Pharma). Participants were instructed to take three tablets daily for 12 weeks.

Outcome measures
The scales used in the questionnaires were devised for this study and based on previously validated self assessment scales.⁸ The questionnaires contained 21 questions about the severity of tinnitus. These were divided into three groups and summary scores were produced for each group. These scores ranged from 0 to 12 for measures of loudness, from 0 to 22 for measures of awareness and ability to ignore, and from 0 to 39 for impact. The sum of the scores in these three groups was the total summary score. A summary score of 0 indicates that a participant has no tinnitus. The maximum summary score of 73 indicates that a participant has tinnitus that is severely troublesomefor example, it is always very loud, the participant can never ignore it, and it has a large impact on the participant's life.

Data analysis
Data were analysed on an intention to treat basis wherever possible. Data entry and initial analyses were carried out by a researcher blinded to the participant's allocation. Statistical analysis was carried out using SPSS version 9.0 for Windows except for the calculation of confidence intervals for proportions.⁹ All reported P values are two tailed. Paired data were compared between treatment groups using McNemar's test and paired sample t tests. Unmatched analyses did not provide any additional information and have therefore been excluded from this paper.

	Results

Top Abstract Introduction Participants and methods Results Discussion References

Altogether 1121 participants were allocated to treatment (559 to active treatment and 562 to placebo) (figure); of these, 956 participants were paired. Characteristics of the paired participants are shown in table 1. Analysis of the side effects of treatment was carried out using data from all 489 matched pairs. However, 26 participants completed no questionnaires so all other analyses were carried out on the remaining 478 pairs in which both members completed at least one questionnaire.

View larger version (39K): [in this window] [in a new window]   Progress of participants through the trial

Outcome measures
There were no significant differences between the treatment groups at weeks 4, 12, and 14 with respect to primary outcome measures (table 2), secondary outcome measures, compliance, or cerebral insufficiency. Additional details are available on the BMJ 's website.

	Discussion

Top Abstract Introduction Participants and methods Results Discussion References

Ginkgo biloba extract LI 1370 had no greater therapeutic effect than placebo in treating tinnitus. In addition, other symptoms of cerebral insufficiency were not significantly affected by the treatment. The results from this trial are similar to some reports and contrast with others.² The main strength of this study was its large size and controlled design. Previous trials involved fewer than 300 subjects and often lacked adequate controls.² This study achieved its large sample size using a simple approach to data collection (postal questionnaires).

Methods of assessing tinnitus have differed between trials, although most have used a simple, subjective measurement of change in tinnitus, similar to the primary outcome measure used in this study. Our method of assessing tinnitus was thorough, enabled small changes to be identified, and concentrated on the most clinically relevant measurement for this condition (that is, perceived changes in tinnitus). Another strength of this study was that this treatment regimen has been shown to be effective in cerebral insufficiency. Additionally, a measure of the symptoms of cerebral insufficiency was included in the design to determine whether any improvements in tinnitus were associated with improvements in symptoms of cerebral insufficiency.

Most previous trials have used similar treatment doses and been of similar duration, but the methods of administration and the composition of the extract have varied.⁵ Therefore, it is possible that at least some of the inconsistencies identified by previous studies may be related to the different types of Ginkgo biloba extract that were used. Measurements of other symptoms of cerebral insufficiency have not been made in previous trials. Since neither tinnitus nor other symptoms of cerebral insufficiency were significantly improved in this study, it would be interesting to learn whether trials in which Ginkgo biloba was found to be effective in tinnitus showed that participants had any improvements in other symptoms of cerebral insufficiency. It is tempting to speculate that positive trials have involved a greater number of patients who have cerebral insufficiency and thus improvements in tinnitus were related to an improvement in cerebral insufficiency rather than being a direct effect of treatment.

The extract used in this study (LI 1370 150 mg/day for 12 weeks) seems to be ineffective in treating tinnitus alone, but it may be effective in treating tinnitus in patients who also have other symptoms of cerebral insufficiency. The composition of other extracts or the use of other treatment regimens, or both, might be effective in treating tinnitus alone but there is little evidence of this.

	Acknowledgments

We thank Mr H Ross for his statistical advice and technical assistance; Dr J Simpson, Mr P Josling, and Dr R Middleton for their help and advice; and members of the Birmingham BTA group and Mr P Hopkins for their administrative help.

Contributors: ED initiated the research. SD and ED designed the study. SD conducted the research and analysed the data. The paper was co-written by the authors. SD is guarantor for the paper.

	Footnotes

Funding: This work was funded by the British Tinnitus Association in conjunction with Lichtwer Pharma UK, manufacturer of the extract used in this study.

Competing interests: The study was financed (two years' salary for SD and running costs) by a contract between the British Tinnitus Association and Lichtwer Pharma GmbH, Berlin, who also supplied the Ginkgo biloba extract and placebo tablets.

The full version of this paper appears on the BMJ's website

	References

Top Abstract Introduction Participants and methods Results Discussion References

1.	Davis A, Rafaie EA. Epidemiology of tinnitus. In: Tyler RS, ed. . Tinnitus handbook. San Diego, CA: Singular Press, 2000.
2.	Ernst E, Stevinson C. Ginkgo biloba for tinnitus: a review. Clin Otolaryngol 1999; 24: 164-167[Medline].
3.	Winslow LC, Kroll DJ. Herbs as medicines. Arch Intern Med 1998; 158: 2192-2199[Medline].
4.	Soholm B. Clinical improvement of memory and other cognitive functions by Ginkgo biloba: review of relevant literature. Adv Ther 1998; 15: 54-65[Medline].
5.	Kleijnen J, Knipschild P. Ginkgo biloba for cerebral insufficiency. Br J Clin Pharmacol 1992; 34: 352-358[Medline].
6.	Blumenthal M, Busse WR, Goldberg A, Gruenwald J, Hall T, Riggins CW, et al, eds. The complete German commission E monographs: therapeutic guide to herbal medicines. Austin, TX: American Botanical Council, 1998.
7.	Duckert LG, Rees TS. Placebo effect in tinnitus management. Otolaryngol Head Neck Surg 1984; 92: 697-699[Medline].
8.	Axelsson A, Coles R, Erlandsson S, Vernon MM, Vernon J. Evaluation of tinnitus treatment: methodological aspects. J Audiological Med 1993; 2: 141-150.
9.	Gardner MJ, Altman DG. Calculating confidence intervals for proportions and their differences. In: Gardner MJ, Altman DG, eds. Statistics with confidence: confidence intervals and statistical guidelines. London: BMJ Publishing, 1989.

(Accepted 11 October 2000)