Trial flow diagram. AMS indicates acute mountain sickness; bid, twice a day.
Dot plot of Lake Louise Acute Mountain Sickness Scoring System (LLS) scores.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Chow T, Browne V, Heileson HL, Wallace D, Anholm J, Green SM. Ginkgo biloba and Acetazolamide Prophylaxis for Acute Mountain Sickness: A Randomized, Placebo-Controlled Trial. Arch Intern Med. 2005;165(3):296–301. doi:10.1001/archinte.165.3.296
Copyright 2005 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2005
Acute mountain sickness (AMS) commonly occurs when unacclimatized individuals ascend to altitudes above 2000 m. Acetazolamide and Ginkgo biloba have both been recommended for AMS prophylaxis; however, there is conflicting evidence regarding the efficacy of Ginkgo biloba use. We performed a randomized, placebo-controlled trial of acetazolamide vs Ginkgo biloba for AMS prophylaxis.
We randomized unacclimatized adults to receive acetazolamide, Ginkgo biloba, or placebo in double-blind fashion and took them to an elevation of 3800 m for 24 hours. We graded AMS symptoms using the Lake Louise Acute Mountain Sickness Scoring System (LLS) and compared the incidence of AMS (defined as LLS score ≥3 and headache).
Fifty-seven subjects completed the trial (20 received acetazolamide; 17, Ginkgo biloba, and 20, placebo). The LLS scores were significantly different between groups; the median score of the acetazolamide group was significantly lower than that of the placebo group (P = .01; effect size, 2; and 95% confidence interval [CI], 0 to 3), unlike that of the Ginkgo biloba group (P = .89; effect size, 0; and 95% CI, −2 to 2). Acute mountain sickness occurred less frequently in the acetazolamide group than in the placebo group (effect size, 30%; 95% CI, 61% to −15%), and the frequency of occurrence was similar between the Ginkgo biloba group and the placebo group (effect size, −5%; 95% CI, −37% to 28%).
In this study, prophylactic acetazolamide therapy decreased the symptoms of AMS and trended toward reducing its incidence. We found no evidence of similar efficacy for Ginkgo biloba.
Acute mountain sickness (AMS) is a common problem among unacclimatized individuals who venture to altitudes above 2000 m.1 Reported incidences from various investigators range from 25% at altitudes of 2000 m2,3 to 53% at those above 4000 m.4 Onset of AMS occurs within a few hours of arrival to high altitude and is determined by the altitude, rate of ascent, and individual susceptibility. Mild AMS, although debilitating, is generally self-limiting and subsides after a period of acclimatization. Persons with moderate or severe symptoms should descend to lower altitudes.1
The most effective preventive measure for AMS—gradual ascent—is frequently difficult or impractical for modern international travel to locations such as Lhasa in Tibet (3650 m) and La Paz in Bolivia (3740 m). Alternatively, acetazolamide therapy4,5 is widely regarded as effective prophylaxis against AMS,1 but it can cause adverse effects and requires a prescription. Ginkgo biloba, an over-the-counter herbal supplement with essentially no adverse effects,6 has demonstrated conflicting results when used for AMS prophylaxis.7-12 Although 4 of the 6 studies of Gingko biloba for AMS prophylaxis that have been conducted thus far have reported efficacy with the use of this supplement, 1 of the 2 studies with negative findings was also the largest. In an attempt to resolve this discrepancy, we compared the effectiveness of Ginkgo biloba use with acetazolamide use in the prevention of AMS in a randomized, placebo-controlled trial.
This randomized, placebo-controlled study was approved by the institutional review board at Loma Linda University Medical Center, Loma Linda, Calif, and all subjects provided written informed consent before entry. We included adult volunteers aged 18 to 65 years whose primary residence was at an elevation of 1200 m or lower. We excluded those who (1) had traveled to an elevation above 2400 m within 30 days of the study; (2) had contraindications to high altitude exposure (eg, uncompensated cardiopulmonary disease, neurologic disorders, or sickle cell disease)13; (3) were pregnant; (4) had preexisting use of acetazolamide or Ginkgo biloba; (5) had a known hypersensitivity to acetazolamide, sulfa medication, or Ginkgo biloba; (6) had known bleeding disorders or were receiving anticoagulant therapy; or (7) had scheduled a surgical or dental procedure within 14 days of study participation.
Each participant was randomly assigned to receive acetazolamide, Ginkgo biloba, or placebo in double-blind fashion. Study medications were prepared as identical-appearing capsules by our hospital pharmacy, placed in identical-appearing packages with enclosed administration instructions, and affixed with serial study numbers. In the event of an emergency, an investigator had access to the study key, which was stored within a sealed envelope. We developed a randomization sequence by drawing cards out of a hat, using 25 labeled cards for each group. Effective randomization for this sequence of 75 numbers was achieved by passing out study drug packages in numerical order.
We obtained Ginkgo biloba over-the-counter from a chain health supplement store. The specific formulation, “Ginkgo Biloba 120 mg Vegetarian” (NOW Foods, www.nowfoods.com), has labeling typical for this agent specifying that it contains a minimum of “24% ginkgoflavonglycosides” and “6% terpene lactones.” We prescribed a “high dosage,” according to the typical use of this herbal agent (120 mg taken orally twice a day [bid]),6 and the subjects began therapy 5 days before their ascent, as has been previously recommended for AMS prophylaxis.8 To maintain blinding, subjects in the acetazolamide group started taking placebo 5 days before ascent and switched to a typical dosage for AMS prophylaxis (250 mg taken orally bid)1 of a standard generic acetazolamide formulation 1 day before ascent.
We instructed the participants to begin taking their study medications 5 days in advance of the ascent, to refrain from travel to elevations above 1200 m for 7 days before the ascent, and to abstain from using alcohol and sedatives during the entire study period. On the evening before the ascent, the participants were driven from the Los Angeles, Calif, metropolitan area to the University of California White Mountain Research Station Owens Valley Laboratory, near Bishop, Calif (elevation, 1230 m). The next morning, they were driven 2 hours up to the University of California Barcroft Laboratory (elevation, 3800 m), where they remained for 24 hours.
To ensure a similar level of physical activity, we required the subjects to abstain from strenuous activity and did not permit them to wander beyond the basic vicinity of the research station. The Barcroft Laboratory is small and could not accommodate all subjects at once. Thus, the participants were divided into 3 separate scheduled ascents, all within the same summer month. Each ascent followed the identical protocol and was directly supervised by the 2 primary investigators (T.C. and V.B.). The investigators continuously monitored the subjects for evidence of high-altitude pulmonary edema (defined as severe cough, dyspnea at rest, cyanosis, and rales) or high-altitude cerebral edema (defined as ataxia, altered consciousness, or both in someone with AMS or high-altitude pulmonary edema).1 Immediate evacuation and treatment with dexamethasone and oxygen were available for such an occurrence.
We graded AMS using the Lake Louise Acute Mountain Sickness Scoring System (LLS), a simple and widely accepted tool for AMS assessment developed by the Lake Louise Consensus Group.14-16 The LLS rates 5 symptoms (headache, gastrointestinal symptoms such as nausea and vomiting, fatigue and/or weakness, dizziness and/or light-headedness, and difficulty sleeping) on a self-report questionnaire and 3 physical examination findings (altered mental status, ataxia, and peripheral edema) on a clinical assessment, with each item graded on a scale from 0 to 3. Thus, the LLS scores can range from 0 (no symptoms or signs) to 24 (worst rating on each scale). As per the Lake Louise Consensus Group, we defined AMS as the presence of headache and at least 1 of the following symptoms: gastrointestinal upset (anorexia, nausea, or vomiting), fatigue or weakness, dizziness or lightheadedness, and difficult sleeping. A score of 3 points or greater on the AMS self-report questionnaire alone, or in combination with the clinical assessment score, constitutes AMS.9,14
The study subjects completed the LLS self-report questionnaire before the ascent, remained at altitude for 24 hours, and then completed the LLS self-report questionnaires before the descent, reflecting maximal symptoms during the study period. Investigators monitored them for the pertinent clinical signs throughout their time at altitude. They were not permitted to self-treat their AMS symptoms. Instead, they were told in advance that investigators would provide oral acetaminophen (1000 mg) or ibuprofen (400 mg) as needed for headache or intramuscular promethazine (50 mg) as needed for nausea and/or vomiting.
Just before their descent, the subjects completed a form on which they circled new symptoms that they had experienced since they started taking the study medication, with the following options: cardiovascular (fainting, heart racing, or heart skipping), dermatologic (hives, itching, or rash), ear/nose/throat (dry mouth, ringing in the ears, runny nose, or sore throat), gastrointestinal (abdominal distention, change in stool color, constipation, or diarrhea), genitourinary (blood in urine, decreased urination, flank pain, increased urination, or pain with urination), neurologic (altered taste or smell, blurred vision, hearing loss, muscular weakness, numbness, paralysis, seizure, tingling in hands/feet/lips/tongue, tremor, or vertigo [spinning sensation]), psychiatric (confusion, depression, drowsiness, excitement, irritability, or nervousness), and other (cough, excessive thirst, fever, joint pain, or weight loss). We also gave them the opportunity to record any symptoms that were not on this list. Finally, also just before the descent, we asked them all to provide their best guess as to whether they had been assigned to the Ginkgo biloba, acetazolamide, or placebo group.
Our primary outcomes were the LLS self-report questionnaire scores and the incidence of AMS. Our secondary outcomes were the number of subjects requesting analgesics, the number requesting antiemetics, the number experiencing high-altitude pulmonary edema or high-altitude cerebral edema, and the incidence of other symptoms.
We used the Kruskall-Wallis, χ2, or Fisher exact test to compare baseline characteristics, as appropriate. We contrasted the distributions of LLS scores with dot plots and compared them using the Kruskall-Wallis test (followed by Wilcoxon rank sum tests if P<.05). We compared the relative incidence of AMS and the secondary outcomes using the χ2 or Fisher exact test, as appropriate.
We performed a sample size calculation based on a control incidence of AMS above 3000 m as 42%.1 To detect a 50% relative reduction from this baseline (α = .05; β = .20) would require 69 subjects in each treatment arm. Unfortunately, we could not achieve this number for logistical reasons and, instead, selected a total sample size of approximately 20 per group to mirror the precedent of other similar research.7-12,17 We recognized in advance that this action would put any negative results at risk for a type II statistical error, and, accordingly, we calculated effect sizes and their confidence intervals (CIs) to quantify this potential error.
We considered P values of less than .05 as demonstrating a statistically significant association and P values of .05 to .10 as demonstrating a trend toward an association. Confidence intervals of the difference between medians were calculated with methods described by Daniel.18 All other analyses were performed using Stata 8 statistical software (Stata Corp, College Station, Tex).
The trial flow diagram is shown in Figure 1. One subject in the placebo group developed high-altitude cerebral edema 10 hours after ascent (details described herein) and was emergently evacuated. She was not removed from the study, however, as she clearly had AMS, and her outcome measures were readily obtained. Her asymptomatic spouse, who was in the acetazolamide group and who descended with her, on the other hand, was removed from the study because his ability to continue 24 hours without AMS could not be determined.
Baseline characteristics were similar between the groups (Table 1), and all subjects reported no symptoms on their LLS self-report questionnaires before the ascent. The subjects guessed their specific drug allocation at a level consistent with chance alone (21 correct of 57 total guesses). All study participants resided in the Los Angeles metropolitan area at elevations above 500 m. No subjects received a study medication other than that to which they were randomized.
Because the LLS scores were significantly different between groups (Figure 2 and Table 2), we used the Wilcoxon rank sum test to compare the LLS scores of both the acetazolamide group and the Ginkgo biloba group with those of the placebo group. The median LLS score of the acetazolamide group was significantly lower than that of the placebo group (P = .01; effect size, 2; 95% CI, 0 to 3), while that of the Ginkgo biloba group was not (P = .89; effect size, 0; 95% CI, −2 to 2).
The incidence of AMS trended toward a difference between groups (Table 2). Acute mountain sickness occurred less frequently in the acetazolamide group than in the placebo group (effect size, 30%; 95% CI, 61% to −15%), and the frequency of occurrence was similar between the Ginkgo biloba and the placebo groups (effect size, −5%; 95% CI, −37% to 28%).
The number of subjects who requested analgesics for headache differed significantly between groups, while the number of those requesting antiemetics for nausea or vomiting did not (Table 2). Treatment for headache was 30% less likely in the acetazolamide group than in the placebo group (95% CI, 61% to −15%) and 4% more likely in the Ginkgo biloba group than in the placebo group (95% CI, 35% to −26%). Treatment for nausea and/or vomiting was 15% less likely in the acetazolamide group than in the placebo group (95% CI, 31% to −1%) and 9% less likely in the Ginkgo biloba group than in the placebo group (95% CI, 29% to −11%).
No subjects developed high-altitude pulmonary edema. The placebo subject who developed high-altitude cerebral edema was a 57-year-old woman who reported headache, nausea, weakness, and dizziness within 6 hours of arriving at altitude. At 10 hours, she developed confusion and required assistance with ambulation, whereupon she was given intramuscular dexamethasone (4 mg) and evacuated to below 1000 m. Magnetic resonance imaging of her brain revealed no abnormalities. During her hospitalization, she received oxygen and dexamethasone, with gradual resolution of her neurologic symptoms. At the 3-month follow-up visit, she demonstrated baseline neurologic function.
The most common symptom of participants (Table 2) in the acetazolamide group was paresthesia. Heart racing, dry mouth, drowsiness, and increased urination were reported in similar proportions by all 3 treatment groups.
This comparative trial of acetazolamide vs Ginkgo biloba for AMS prophylaxis confirmed a beneficial effect for acetazolamide but failed to demonstrate even a trend to support the efficacy of Ginkgo biloba therapy. Although our study was not powered to definitively exclude the benefit of Ginkgo biloba use, the 95% CIs of our effect sizes demonstrate that at most we might have missed an improvement of 2 points in the LLS score and a 28% decrease in the incidence of AMS. These data do not support the use of Ginkgo biloba as a replacement for acetazolamide therapy.
Acetazolamide is widely recommended for AMS prophylaxis,1,4,5,13 and our results confirm its efficacy for this indication. The primary disadvantages of the use of acetazolamide for AMS prophylaxis are the need for a prescription and complicating paresthesias (noted in 7 of 20 subjects). Other adverse effects (eg, polyuria and altered taste) occurred at rates similar to those induced by placebo.
In the United States, Ginkgo biloba is classified as an herbal supplement and is available without a prescription. It is promoted for the treatment of dementia, claudication, and tinnitus. Its only established adverse effect with potentially serious consequences is platelet inhibition with increased bleeding risk.6,19 It has been studied for AMS prophylaxis in 6 trials (4 with positive results; 2 with negative results), 3 of which have been published thus far only in abstract form.8,10,12
In the first of the 4 trials with positive results, Roncin et al7 compared Ginkgo biloba extract (80 mg bid) with placebo for the prevention of AMS in 44 subjects during a gradual (8-day) Himalayan ascent to 4900 m and graded AMS symptoms using the cerebral and respiratory scales of the Environmental Symptoms Questionnaire. Using the cerebral scale, none of 22 subjects in the Ginkgo biloba group developed AMS compared with 9 of 22 subjects in the placebo group; while using the respiratory scale, AMS was noted in 3 of 22 subjects in the Ginkgo biloba group compared with 18 of 22 subjects in the placebo group (both groups, P≤.001). The study does not state whether it used blinding.
Maakestad et al8 report in abstract form a randomized, double-blind trial of Ginkgo biloba (120 mg bid starting 5 days before ascent) vs placebo for the prevention of AMS in 40 college students undergoing rapid ascent from 1400 to 4300 m. Using the LLS and Environmental Symptoms Questionnaire as outcomes, AMS was noted in 7 of 21 subjects in the Ginkgo biloba group compared with 13 of 19 subjects in the placebo group (P<.02). The abstract does not state whether the trial was randomized.
Gertsch et al9 performed a randomized, double-blind trial of Ginkgo biloba (60 mg 3 times a day starting 1 day before ascent) vs placebo for the prevention of AMS in 26 subjects who ascended rapidly to 4205 m. They noted a trend (P = .07) toward efficacy; AMS (defined as an LLS score ≥3) developed in 7 of 12 subjects in the Ginkgo biloba group compared with 13 of 14 subjects in the placebo group. In the last of the 4 trials with positive results, Moraga et al10 report in abstract form a comparison of prophylaxis with Ginkgo biloba (80 mg bid) vs acetazolamide (250 mg bid) vs placebo that was started 24 hours before rapid ascent to 3700 m. Randomization and blinding are not described. Of 32 subjects enrolled, none of those in the Ginkgo biloba group developed AMS compared with 35% of those in the acetazolamide group and 54% of those in the placebo group. The 2 trials with negative results were both experiments that had been conducted previously by investigators with positive results, as reported above. Gertsch et al11 performed a randomized, double-blind, placebo-controlled trial of prophylactic therapy with Ginkgo biloba (120 mg bid), acetazolamide (250 mg bid), both Ginkgo biloba and acetazolamide, or neither that was begun 36 to 48 hours before a Himalayan ascent starting at either 4248 m or 4397 m and ending at 4950 m. Among 487 subjects who completed the trial, AMS was noted in 35% of those who received Ginkgo biloba, 12% of those who received acetazolamide, 14% of those who received both Ginkgo biloba and acetazolamide, and 34% of those who received only placebo. Leadbetter and Hackett12 performed a randomized, double-blind trial in which 59 subjects received Ginkgo biloba (120 bid), acetazolamide (125 mg bid), or placebo starting 3 days before a rapid ascent to 4300 m. Although they confirmed benefit from acetazolamide compared with placebo, they found no statistically significant improvement from Ginkgo biloba. The relative incidences of AMS in the placebo and acetazolamide groups were not reported.
As in the latter 2 studies, our trial found no evidence of efficacy for Ginkgo biloba. Although it is possible that our randomized, double-blind trial missed a true effect as a result of chance alone, our data do not demonstrate even a trend in support of Ginkgo biloba. Indeed, we conservatively designed our trial to avoid any possible bias against Ginkgo biloba: we used a high dose of this agent, a longer period of drug preloading, and a steep ascent profile from sea level. Given our data and those of the 2 prior trials with negative results, it appears that the efficacy of the use of Ginkgo biloba for the prevention AMS can be seriously questioned. Possible explanations for the discrepant outcomes between the various trials may include dosage, duration of therapy before ascent, and speed of ascent. An alternative explanation for the differences in efficacy noted in these trials may be variation in Ginkgo biloba quality or potency. As an herbal supplement, Ginkgo biloba is free from the rigorous regulation to which the United States subjects drugs such as acetazolamide, and there may be considerable variation in the composition between manufacturers and lots.20 Indeed, a recent study of the popular herbal agent Echinacea found that less than 50% of the sampled brands met the quality standard described on the label, and 10% contained no Echinacea at all.21 A lack of bioequivalence has also been noted between brands of Ginkgo biloba.22
Gertsch et al9 specifically described the source of their extract as “packaged in Europe by Boehringer-Ingelheim pharmaceuticals”; however, the Ginkgo biloba sources in the other reports are not clear.8 We intentionally selected an over-the-counter source for our investigation because of the premise that one of the advantages of using Ginkgo biloba is that it is readily available without prescription. We obtained the “standardized” form of Ginkgo biloba that contained the same components in the same proportions as the Ginkgo biloba used in other studies and in accordance with the German E Commission.23 It is possible that the Ginkgo biloba used in our study was less potent than that used in the other trials because of imprecise preparation by its manufacturer. However, we selected a common brand from a reputable health supplement store, and our method would thus mirror the actions of most individuals purchasing this agent for AMS prophylaxis. The primary limitation of our study is its relatively small sample size, a limitation typical of most AMS research, owing to logistical reasons. We do not believe that blinding represents a limitation in our trial, as our subjects were unable to discern their study drug at a rate better than chance alone. Other authors have suggested that the use of acetazolamide cannot be effectively blinded because of polyuria7; however, this symptom was not specific to the subjects who received acetazolamide at the dosage used in our trial.
In conclusion, in our study, prophylactic acetazolamide therapy decreased the symptoms of AMS and trended toward reducing its incidence. We found no evidence of similar efficacy for Ginkgo biloba.
Correspondence: Tony Chow, MD, Department of Emergency Medicine, Loma Linda University Medical Center A-108, 11234 Anderson St, Loma Linda, CA 92354 (Drtkchow@aol.com).
Accepted for Publication: July 31, 2004.
Previous Presentation: This study was presented in part at the 49th Annual Meeting of the American College of Sports Medicine; June 1, 2002; St Louis, Mo.
Financial Disclosure: None.
Create a personal account or sign in to: