Original Research

Effects of Banisteriopsis caapi Extract on Parkinson’s Disease

Abstract. A substance extracted from the Amazonian vine Banisteriopsis caapi was shown in the 1920s to alleviate parkinsonism. These pioneering studies were criticized and forgotten for a number of reasons, including questions as to the identity of the active agent and failure to conduct strictly controlled studies. We now report the first double-blind, randomized placebo-controlled trial of a Banisteriopsis caapi (BC) extract for treatment of Parkinson’s disease (PD). A single dose of BC administered to de novo PD patients resulted in significant improvement in motor function evidenced by decline in the Unified Parkinson’s Disease Rating Scale (UPDRS) score. The beneficial effects were maximal by the second hour and persisted until the last evaluation of the patients at 4 hours. However, tremor was not improved and in some patients tremor was exacerbated. All patients also experienced a degree of transient nausea or vomiting. We measured the concentrations of the putative active agents (harmine, harmaline, and tetrahydroharmaline), and hypothesize that the beneficial effects were primarily due to glutamate receptor antagonist actions of the harmalines.


Ayahuasca is a beverage prepared from a Banisteriopsis caapi (BC) and the leaves of other plants, most commonly Psychotria viridis, which contains the hallucinogen dimethyltryptamine (DMT). Ayahuasca (known as hoasca in Brazil, Yage in Colombia, and Caapi in other parts of the Amazon basin), has a long history of use in medico-magical-religious ceremonies by natives of the Amazonian rain forests.1,2 More recently, ayahuasca has been used as a religious sacrament by two state-approved churches in Brazil (Uniao do Vegetal and Church of Santo Daime). The banisteriopsis vine contains contains b-carbolines such as harmine, harmaline, and tetrahydroharmine,3,4,5 which normally do not produce the desired psychotropic effects unless they are used with the DMT-containing plant. When taken orally, DMT is metabolized by monoamine oxidase (MAO) and is inactive, but the combination of DMT and the MAO inhibitor banisterine results in psychotropic effects.

The active agent of the banisteriopsis vine was identified by Louis Lewin, who named it banisterine. Based on the subjective sensations experienced by Lewin following self-adminstration, he recommended it for treatment of rigid-akinetic syndromes in the late 1920s in Germany;4 all the reports mention improvement in rigidity, but conflicting results regarding tremor. Rustige also noted improvement in mood and affect. Later, Halpern studied the subjective effects on herself, and noted a sensation of lightness of body and a belligerent, aggressive mood. From the 1930s to the present, the use of banisterine in treatment of PD has been virtually forgotten.

One of us (MS-D) had previously evaluated 13 patients who admitted use of ayahuasca occasionally (every 5 or 6 months), and who felt it helped alleviate the signs and symptoms of PD. In 4 of these patients we were able to directly observe the effects of consumption of ayahuasca. In one of these, extract of Banisteriopsis caapi (BC) was taken alone (no admixture of plants) and in the other 3 patients, the admixture of plants was ingested. Approximately 20 minutes after ingestion of the beverage, there was an improvement in rigidity with exacerbation of tremor and appearance of abnormal involuntary movements as well as the induction of a hallucinatory state. The motor effects and hallucinations were much less evident in the single patient who took only the BC extract.

L-dopa is the gold standard of PD treatment, but its use is limited by the development of motor complications, including dyskinesias, in 30 to 80% of chronically treated patients.6 As a result, additional therapies have been developed that increase and stabilize synaptic levels of dopamine without augmenting the administered dosage of L-dopa, as for example by inhibiting dopamine metabolism with MAO inhibitors or with catechol-O-methyl transferase inhibitors.7,8

Based on the remarkable effects of banisterine reported in the late 1920s and the dramatic effects noted in the few patients described above, we undertook a double-blind, randomized, placebo-controlled study of the acute effects of a single dose of BC extract in a cohort of 30 de novo PD patients.

Figure 1


Thirty consecutive patients with newly diagnosed PD were selected (using diagnostic criteria of Calne et al.9) from the Neurology Service of Carlos Andrade Marín Hospital in Quito, Ecuador. The patients were randomly assigned to receive either banisteriopsis extract or placebo tea (Figure 1).

All patients underwent assessment of higher cortical functions with the Short Test of Mental Status (STMS);10 the Hamilton test for depression (HT);11 cranial CT; serum chemistry (urea, creatinine, and glucose); hepatic enzymes (GOT, GPT, GGT); and cardiac evaluation including EKG.

Patients who scored more than 50% on the HT, who scored less than 29 points on the STMS, or who exhibited anomalies on the serum tests or cardiac evaluation were excluded from the study.

BC was prepared as a liquid extract 12 hours before administration in a standard manner: 50 g (dry weight) of the vine was chopped into small fragments, and added to 1500 mL of water, brought to boiling for four hours until the volume was reduced to 200 mL. The solution was filtered and adminstered in that form.

All patients came to the clinic fasting and received 200 mL of either the BC solution (BC) or placebo (P), which consisted of 200 mL of manzanilla tea. Patients were not allowed to eat or drink anything but water during the course of the session.

Patients were evaluated using the motor component of the Unified Parkinson’s Disease Rating Scale (UPDRS)12 before (baseline) and 60, 120, and 240 minutes after ingesting the test infusions (questions 3, 4, 9, 11, 14, 17, 32–39, and 41 of the UPDRS were not included). Analysis of variance (ANOVA) of the data followed by Bonferroni corrections for multiple comparisons was performed. An value of 0.05 was chosen and p < 0.0125 was considered statistically significant.13

This protocol was approved by the Department of Docencia and Research of the Andrade Marín Hospital in Quito, Ecuador. All patients provided informed consent.

Measurement of harmala alkaloids in Banisteriopsis caapi extract

Analysis of the levels of harmala alkaloids (harmine, harmaline, and tetrahydroharmine) present in the banisteriopsis extract was performed as described previously3 with certain modifications. HPLC system: The mobile phase was 100 mM ammonium acetate, pH 6.9, with 20% methanol and 20% acetonitrile. Flow rate was kept at 1 mL/min using a BAS 200 series pump (BAS, W. Lafayette, Ind, USA). The mixture of harmala alkaloids was separated using a reverse phase YMCbasic HPLC column (4.6 × 150 mm) with a 3-micron particle size (YMC Inc., Wilmington, NC, USA). Harmine and harmaline were purchased from Sigma Chemical Co. (St. Louis, Mo, USA). Tetrahydroharmine was produced from the reduction of harmaline with sodium borohydride, as described previously.3 Harmine, harmaline, and tetrahydroharmine were detected using a LS-3B fluorescent detector (Perkin Elmer, Norwalk, Conn, USA). Harmine was detected using a wavelength of 300 nm for excitation and 360 nm for emission. Harmaline and tetrahydroharmine were detected with wavelengths of 340 nm and 480 nm for excitation and emission, respectively. A calibration curve was constructed with known amounts of the 3 compounds and the area of the peaks was recorded and compared to the areas produced by the injection of the Banisteriopsis caapi extract. Levels of harmala alkaloids were expressed as micrograms per milliliter of extract.


Table 1
The average age of the BC group was 67 ± 8.3 years and the P group averaged 62.9 ± 7 years of age. Median duration of disease was 24.8 ± 12 months in the BC group and 25 ± 7.7 months in the P group. Fifteen patients were male and 15 female (Table 1).

At baseline, the mean UPDRS score in the BC group was 54.4, and 53.8 for the P group (p < 0.4). The UPDRS scores at 60, 120, and 240 minutes were 41.4 in the BC group compared to 46.4 in the P group; 22.4 in the BC group compared to 52.5 in the P group; and 25.6 in the BC group compared to 53 in the P group, respectively. These differences were statistically significant (p < 0.0004; p < 0.0001; p < 0.0001) (Table 2).

Table 3

Side effects were noted only in the BC group. Nausea or vomiting was noted in 100% of the BC patients, diarrhea in 53.3%, agitation in 26.6%, and 1 patient experienced transient hallucinations. All of the patients in the BC group experienced an exacerbation of their resting tremor and, in addition, postural and action tremor. Abnormal involuntary movements, choreiform in nature, were also noted (Table 3).

Analysis of the Banisteriopsis caapi extract yielded the following levels of harmala alkaloids (mean ± sem): harmine 418.44 ± 11.78 m g/mL, harmaline 173.03 ± 4.13 m g/mL, and tetrahydroharmine 382.40 ± 8.38 m g/ml.

Table 2


In this double-blind, randomized, placebo-controlled trial, we demonstrated that a single dose of BC administered to de novo PD patients resulted in significant improvement in motor function evidenced by decline in the motor component of the UPDRS score. The beneficial effects were noted by 1 hour and motor function continued to improve for the 4 hours during which the patients were studied. However, all patients who received BC experienced a worsening of resting tremor and the development of action and postural tremors, with some abnormal choreiform movements. All patients also experienced a degree of transient nausea or vomiting. With the exception of the single patient who experienced confusion and hallucinations, these side effects were much less severe than those experienced by users of the complete ayahuasca beverage.

The levels of harmaline in the banisteriopsis extract ingested by the subjects in this study were almost identical to those measured by Callaway and colleagues in the preparations of ayahusaca used by members of the Uniao de Vegetal in Brazil.3 However, the mean harmine and tetrahydroharmine levels were 25% and 35%, respectively, of the levels measured in their Brazilian subjects, who ingested the complete brew prepared from a combination of plants.3 The proportion of banisteriopsis vine to Psychtria viridis leaves and the exact methods for preparation of the brew were not noted in the Callaway paper.

In the shamanic use of ayahuasca for religious, magical, or healing purposes, the infusion is prepared from a mixture of plants. A critical component of the shamanic beverage is the presence of the DMT-containing Psychotria viridis.1,2,3 DMT is a potent hallucinogen when given systemically or when smoked; when taken orally, it is inactive due to its metabolism by gut and liver MAO. BC is an inhibitor of MAO and, when ingested orally with DMT-containing plants, allows the DMT to produce a range of psychotropic effects, with prominent visions, hallucinations, and illusions.1,2,5

The dramatic improvement in signs and symptoms of PD produced by the extract of BC may be due to a combination of two known mechanisms of action of the harmalines, the putative active agents. Harmaline, a bcarboline compound with a structural resemblance to serotonin, is a known nonselective inhibitor of MAO.14 MAO inhibitors can potentiate the actions of endogenous dopamine, but the symptomatic benefits in PD are very mild, as evidenced by the study of deprenyl in denovo patients.15 It is difficult to conceive that the dramatic improvement produced by BC can be due solely to MAO inhibition. It is possible that the interaction of harmaline at glutamatergic receptors plays a significant role in restoring motor function in PD. Harmaline has been shown to be an NMDA receptor antagonist.16 Harmaline displaces [3H]JMK-801, which binds to the cation channel of the NMDA receptor, from membranes prepared from rabbit brain tissue.16

Glutamate is an excitatory amino acid that plays a role in the symptomatic expression of PD17 and that has also been implicated in the process of neurodegeneration of dopaminergic neurons of the substantia nigra.18 When dopamine deficiency develops, the adaptive changes in the striatal outflow pathways result in disinhibition of the subthalamic nucleus. In turn, the hyperactive subthalamic-pallidal glutamate projection results in decreased outflow from globus pallidus internal segment (GP-int) to thalamus, to produce the clinical manifestations of slowness and rigidity.19 Blockage of glutamatergic receptors corrects the imbalance that results from dopamine deficiency and helps restore normal motor function20 In the rat, local infusion of NMDA receptor antagonists to the GP-int/SNr has been shown to reverse the signs of parkinsonism.21 In both rats and primates, the NMDA receptor antagonist LY235959 has been shown to potentiate the antiparkinson effects of L-dopa, to stabilize the motor fluctuations and to alleviate choreiform dyskinesias. However, MK-801, a glutamate NMDA receptor antagonist with hallucinogenic side effects (similar to phencyclidine and ketamine) has been reported to induce dystonia when used with L-dopa in a primate model of parkinsonism.21 In native rats, MK-801 increases locomotor activity and potentiates the motor effects of L-dopa.23 The AMPA antagonist NBZX (2,3-dihydroxy-6-intro-7-sulfamoil-benzo-9[f]quinoxaline) improved tremor, posture, and manual dexterity in parkinsonian monkeys.24 When glutamate antagonists are given together with L-dopa, the dosage of L-dopa can be greatly reduced without lessening motor function.20,24 To conclude, we suggest that the antiglutamate actions of harmaline are most likely responsible for the antiparkinson effects observed in the BC group of patients.

The transient worsening of tremor and induction of postural and action tremor by banisterine has been noted before by early investigators and, in fact, harmaline was used in the 1960s to study the mechanism of action tremor that could be induced by harmaline in primates.25

Interestingly, L-dopa-induced dyskinesias are associated with a decrease in the activity of the glutamatergic projection to the GP-int. Paradoxically, glutamatergic receptors are apparently involved in these dyskinesias, since adminstration of glutamate antagonists (of the NMDA receptor subtype) have been shown to be useful in controlling L-dopa-induced dyskinesias.20

There is little concern for the possibility that the harmalines in the ayahuasca tea are cytotoxic at concentrations found in the tea. Harmaline and a related bcarboline, ibogaine, are cytotoxic in rats only following extremely high doses of 100 mg/kg or 100 mg/kg × 3. Two recent studies clearly demonstrated that even lower doses (25 and 40 mg/kg) given to rats did not produce Purkinje cell degeneration.25,26

To summarize, extracts of BC were shown to have remarkable symptomatic benefit in drug-naive de novo patients. This was the first double-blind, randomized, placebo-controlled trial of a BC extract for treatment of PD. Studies with banisterine done in the late 1920s were criticized because the benefits were believed by many to be psychological. We measured the concentrations of the putative active agents, and hypothesize that the beneficial effects were primarily due to glutamate receptor antagonist actions of the harmalines. The most common side effect was transient nausea/vomiting, which could in the future be prevented by pretreatment with domperidone. Hallucinations in a single case may reflect a higher sensitivity in that patient to the NMDA receptor antagonists. A complete dose-response study needs to be done in human volunteers in order to determine the optimal concentrations of b-carbolines to produce maximal symptomatic relief and minimal side effects.


  1. Ayala Flores F, Lewis WH. Drinking the South American hallucinogenic ayahuasca. Econ Botany. 1978; 32: 154–156.
  2. Bennett BC. Hallucinogenic plants of the Shuar and related indigenous groups in Amazonian Ecuador and Peru. Brittonia. 1992; 44: 483–493.
  3. Callaway JC, Raymon LP, Hearn WL, et al. Quantitation of N,N-dimethyltryptamine and Harmala Alkaloids in Human Plasma after Oral Dosing with Ayahuasca. J Anal Toxicol. 1996, 20: 492–498.
  4. Sanchez-Ramos JR. Banisterine and Parkinson’s disease. Clin Neuropharmacol. 1991; 14: 391–402.
  5. McKenna DJ, Towers GH, Abbott F. Monoamine oxidase inhibitors in South American hallucinogenic plants: tryptamine and beta-carboline constituents of ayahuasca. J Ethnopharmacol. 1984: 10; 195–223.
  6. Bédard PJ, Blanchet PJ, Lévesque D, et al. Pathophysiology of L-dopa-induced dyskinesias. Mov Disord. 1999; 14(Suppl 1): s4–s8.
  7. Lang AE, Lozano AM. Parkinson’s disease. First of two parts. N Engl J Med. 1998; 339: 1044–1053.
  8. Lang AE, Lozano AM. Parkinson’s disease. Second of two parts. N Engl J Med. 1998; 339: 1130–1143
  9. Calne DB, Snow DB, Lee C. Criteria for diagnosis in Parkinson’s disease. Ann Neurol. 1992; 32: 125–127.
  10. Kokmen E, Naessens JM, Offord KP. A Short test of mental status: description and preliminary results. Mayo Clin Proc. 1987; 62: 281–288.
  11. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960; 23: 56–62.
  12. Fahn S, Elton RL, and members of The UPDRS development committee. Unified Parkinson’s disease rating scale. In: Fahn S, Marsden CD, Goldstein M, Calne DB, eds. Recent Development in Parkinson’s Disease. Vol. 2. Florham Park, NJ: MacMillan; 1987: 153–163.
  13. Dawson-Saunders B, Trapp RG. Bioestadística Médica. México DF: Editorial El Manual Moderno, 1993 [Basic and Clinical Biostatics. Trans. Carsolio M. New York, NY: Appleton & Lange; 1990].
  14. Udenfriend S, Witkop B, Redfield BG, Weissbach H. Studies with reversible inhibitors of monoamine oxidase: harmaline and related compounds. Biochem Pharmacol. 1958; 1: 160–165
  15. Parkinson Study Group Effects of tocopherol and deprenyl on the progression of disability in early Parkinson’s disease. N Engl J Med. 1993; 328: 176–183.
  16. Du W, Aloyo VJ, Harvey JA. Harmaline competitively inhibits [3H]MK-801 binding to the NMDA receptor in rabbit brain. Brain Res. 1997; 770(1–2): 26–29.
  17. Ciliax BJ, Greenamyre T, Levey AI. Functional, biochemistry and molecular neuropharmacology of the basal ganglia and motor systems. In: Watts RL, Koller WC. Movement Disorders. Neurologic Principles and Practice. New York, NY: McGraw-Hill; 1997: 99–116.
  18. Olanow CW, Jenner P, Tatton NA, Tatton WG. Neurodegeneration and Parkinson’s disease. In: Jankovic J, Tolosa E. Parkinson’s Disease and Movement Disorders. 3d ed. Baltimore, Md: Williams & Kilkins; 1998: 67–103.
  19. Young AB, Penney JB Jr. Biochemical and functional organization of the basal ganglia. In: Jankovic J, Tolosa E. Parkinson’s Disease and Movement Disorders. 3rd ed. Baltimore, Md: Williams & Kilkins; 1998: 1–13.
  20. Blandini F, Greenamyre JT. Prospects of glutamate antagonists in the therapy of Parkinson’s disease. Fundam Clin Pharmacol. 1998; 12: 4–12.
  21. Ossowska K, Lorenc-Koci E, Konieczny J, Wolfarth S. The role of striatal glutamate receptors in models of Parkinson’s disease. Amino Acids. 1998; 14: 11–15.
  22. Starr MS. Antagonists of glutamate in the treatment of Parkinson’s disease: from laboratory to the clinic. Amino Acids. 1998; 14: 41–42.
  23. Liljequist S, Ossowska K, Grabowska-Anden M, Anden NE. Effect of the NMDA receptor antagonist, MK-801, on locomotor activity and on the metabolism of dopamine in various brain areas of mice. Eur J Pharmacol. 1991; 195: 55–61.
  24. Lange KW, Reiderer P. Glutamatergic drugs in Parkinson’s disease. Life Sciences. 1994; 55: 2067–2075.
  25. Poirier LJ. The development of animal models for studies in Parkinson’s disease. In: McDowell FH, Markham CH, eds. Recent Advances in Parkinson’s Disease. Philadelphia, Pa: F. A. Davis Co; 1971: 83–118.
  26. Molinari HH, Maisonneuve IM, Glick SD. Ibogaine neurotoxicity: a re-evaluation. Brain Res. 1996; 737(1–2): 255–262.
  27. Xu Z, Chang LW, Slikker W Jr, Ali SF, Rountree RL, Scallet AC. A dose-response study of ibogaine-induced neuropathology in the rat cerebellum. Toxicol Sci. 2000; 57(1): 95–101.