Townsend Letter for Doctors and Patients
Alternative Medicine Conference Calendar
Who are we?New articlesFeatured topicsArticles onlineSubscriptionsContact us!
Check out recent tables of contents
From the Townsend Letter for Doctors & Patients
December 2002
Phytotherapy Review & Commentary
Gymnema: A Key Herb in the Management of Diabetes

by Kerry Bone, FNIMH, FNHAA
Our December 2002 cover
Order back issues
Advertise with TLDP!
Order this issue!
Search our site

General Information
      Gymnema sylvestre is a liana or climbing plant with stems up to 8 m in length. It grows in open woods and bushland at an altitude of 100-1000 m in India, China, Indonesia, Japan, Malaysia, Sri Lanka, Vietnam and South Africa. Both the leaf and root are used in Ayurvedic medicine. On account of its property of abolishing the taste of sugar it was given the Hindi names of Gurmar and Madhunashini meaning 'sugar destroying.' The sweet taste suppressant property of Gymnema was revealed to a British officer by the inhabitants of a northern Indian village in the mid-19th century. The herb is traditionally used for the treatment of diabetes and Gymnema extracts are sold in Japan for the control of obesity.

Visit our pre-2001 archives

Clinical Summary

Antidiabetic, hypoglycaemic, hypocholesterolaemic, antiobesity.

Therapeutic Indications
• Hyperglycaemia, insulin-dependent and non-insulin-dependent diabetes mellitus (prolonged administration required).

Standard Process

• Reduction of sweet cravings and appetite.
• Weight loss and dieting since Gymnema (in liquid form) anaesthetises the sweet taste buds.
• May be of benefit for hyperlipidaemia.

Dosage & Administration
      The dosage used in Ayurvedic medicine ranges from 6 to 60 g of dry or powdered leaf per day as an infusion.1-3 The adult dosage for Gymnema 1:1 liquid extract is 25 to 75 mL per week. Some cases of diabetes will respond quickly but best results come after 6 to 12 months of continuous use. Gymnema is conveniently prescribed in tablet form, in this case 8 to 12 g per day of leaf equivalent is recommended. Since they are best enterically-coated, tablets should be taken whole and not crushed.

      For sweet-craving and sweet taste depression 1 to 2 mL per day of liquid extract is all that is necessary. This should be applied in divided doses directly to the tongue by dropper and rinsed off after one minute. This can be done at 2- to 3-hour intervals.

Suggested Combinations
      Gymnema combines well with fenugreek, goat's rue or neem leaf for diabetes, and with globe artichoke or blue flag for weight loss. For hypercholesterolaemia consider combining Gymnema with turmeric, hawthorn, Silybum, globe artichoke or garlic.

Adverse Reactions
      As with all saponin-containing herbs, oral use may cause irritation of the gastric mucous membranes and reflux. To avoid the likelihood of this side effect, enteric-coated tablets are advised.

Contraindications & Cautions: None known.

Traditional Uses

      Gymnema has been described in Hindu Materia Medica as a stomachic, diuretic and antiperiodic (e.g. to treat a periodic illness such as malaria). In Ayurvedic medicine indications for Gymnema include glycosuria, urinary disorders and diabetes mellitus.2 Gymnema has been used in the treatment of diabetes mellitus in India for over 2000 years.4

Scientific Studies

      Gymnema leaf contains 4 to 10% of a group of more than 20 saponin glycosides of the oleanane-type including gymnemic acids I-XVIII and gymnemasaponins I-V.5-7 Some of the gymnemic acids are acylated (contain an acyl group), while the gymnemasaponins are non-acylated. Acylation affects pharmacological activity.

      Dammarane-type saponins8 and a polypeptide consisting of 35 amino acid residues called gurmarin and other oleanane-type saponins (gymnemasins) have also been isolated.

      'Gymnemic acid' referred to in the literature is often not chemically defined and most likely refers to the crude saponin fraction of Gymnema (i.e. all the saponins) or to a mixture of the gymnemic acids.

Sweet Taste Suppression (Antisweet Activity)
      The antisweet principles of Gymnema include gymnemic acids,11 gymnemasaponins6 and gurmarin.12 Gymnemasaponins completely inhibited the perception of sweetness induced by a 0.1 M sucrose solution. This is about half the activity exhibited by the gymnemic acids. (Hence the structure of the saponin affects the magnitude of the sweetness inhibition. Acylation of the saponin, as occurs (naturally) in the gymnemic acids, produces greater inhibition.)6 The reduced sensitivity to sweet substances produced by Gymnema might result from the competition at the receptor sites between glycosides and the sweet substances.13 An electrophysiological study on taste responses in rats suggests that gurmarin acts on the apical side of the taste cell, possibly by binding to the sweet taste receptor protein.14

      In humans and chimpanzees, gymnemic acids suppress the sweet taste of all sweeteners but had no effect on taste receptors sensitive to bitter or salty substances in chimpanzees.15 The sweet suppressing activity of Gymnema extract in humans has been verified by measurement of gustatory evoked potentials.16 In experimental models the degree of suppression produced by gymnemic acid (measured by nerve response) varied from complete abolishment (aspartame, saccharin) to about 50% reduction (xylitol). Gymnemic acid had no effect on the responses to nonsweet (bitter, salty or sour) compounds. These results parallel psychophysical and electrophysical findings in humans.17

      A gymnemic acid rinse used by human volunteers reduced the intensities of sucrose and aspartame to 14% of their pre-rinse levels.18 Over a recovery interval of 30 minutes these values increased linearly to 63% of the pre-rinse levels. A study involving human volunteers observed that a period of at least 30 seconds was required after tasting Gymnema infusion for the full sweet suppression effect to appear.19

      Pretreatment with Gymnema extracts reduced the sweetness of sweeteners by an average of 77% in volunteers. There was no evidence for a differential effect across the range of sweeteners (3 concentrations each of acesulfame K, aspartame, sodium cyclamate, fructose, glucose, sucrose, stevioside and xylitol). The percentage reduction in sweetness was constant across the low, medium and high concentrations. Such results suggest that a receptor occupancy or blocking mechanism is unlikely. A type of 'mixed' inhibition involving an effect on the breakdown of the stimulus/receptor complex is more likely.20

      For further antisweet activity demonstrated in humans see below in Clinical Studies (Weight Loss).21

Hypoglycaemic Activity
      Pharmacodynamic and clinical studies suggest that the hypoglycaemic activity of Gymnema may be mediated through stimulation of insulin release (and possibly by pancreatic regeneration or repair), stimulation of enzymes responsible for glucose uptake and utilisation and/or inhibition of intestinal absorption of glucose.22-25

      Gymnema extract and some isolated constituents have inhibited glucose uptake in isolated small intestinal tissue.26,27 Gymnema extract and gymnemic acids inhibited the intestinal absorption of glucose in humans and rats.28,29 In evidence of this, two fractions obtained from Gymnema (containing gymnemic acids) suppressed potassium-induced contraction of isolated ileal longitudinal muscle, interfered with the increase in transmural potential difference induced by glucose and inhibited the elevation of blood glucose in vivo (route unknown).30 Two gymnemic acids suppressed the contraction of smooth intestinal muscle, most likely by inhibiting glucose uptake.28

      Oral administration of Gymnema extract reduced post-prandial serum glucose and improved glucose tolerance in mildly diabetic rats. Pancreas weight and content of insulin were not changed.31 Gymnema corrected hyperglycaemia in mild alloxan-diabetic rats and significantly prolonged lifespan in severe alloxan-diabetic rats (i.e. with completely destroyed pancreatic tissue). The authors suggested that the prolonged survival time was due to the adaptogenic activity of Gymnema.32

      Feeding with Gymnema leaf powder regulated blood sugar levels in alloxan-diabetic rabbits and increased the activities of enzymes which facilitate the use of glucose by insulin-dependent pathways (phosphorylase, gluconeogenic enzymes and sorbitol dehydrogenase). Uptake of glucose into glycogen was increased in liver, kidney and muscle. Gymnema treatment also increased the incorporation of glucose into the protein components of these tissues.22 The hypoglycaemic activity of Gymnema occurs in a slow and steady manner: in rabbits with mild alloxan-diabetes between 12 and 24 weeks of treatment was required.4

      Blood sugar levels in normal and diabetic rats were lowered 2 hours after oral administration of a Gymnema concentrate (50-400 mg/kg, 19.5:1), but Gymnema was 6 times less potent than oral tolbutamide in the diabetic animals.33 Two Gymnema extracts returned fasting blood glucose levels to normal after 20 to 60 days of oral administration to diabetic rats. A rise in serum insulin towards normal fasting levels occurred and the number of beta cells within pancreatic tissue increased.23 This suggests a restorative effect on pancreatic tissue.

      The crude saponin fraction of Gymnema and gymnemic acid IV reduced blood glucose levels in streptozotocin-diabetic mice when administered by injection. Gymnemic acid IV also increased plasma insulin levels.34 Gymnemoside b and gymnemic acids III, V and VII (route unknown) produced some inhibitory activity on glucose absorption after oral glucose loading in rats, but gymnemic acid I and gymnemasaponin V were inactive.35

Appetite & Weight Loss
      The effect of Gymnema on body weight, glucose absorption and lipid metabolism was examined by using a breed of fatty rats with genetic obese-hyperglycaemia. Simultaneous feeding with Gymnema aqueous extract decreased body weight in fatty and lean rats (4.2% and 6.1% respectively) compared to animals consuming only the test diet (high carbohydrate-low fat) over a 21-week period. Plasma glucose was lower by 18% in the Gymnema-treated animals compared to controls. The plasma glucose increase following an oral glucose tolerance test was almost normalised in the Gymnema-treated group without any alteration in serum insulin levels. Hypertriglyceridaemia, but not hypercholesterolaemia, was also improved in the treated group.36

      Two fractions of Gymnema extract (containing 160-360 mg/g of gymnemagenin) decreased body weight gain and food intake dose-dependently when given orally (0.05-1.0 g/kg) to rats for 22 days. Administration of a Gymnema fraction (1.0 g/kg) containing 363 mg/g of gymnemagenin increased faecal excretion of cholesterol, total neutral steroids, total bile acids and cholic acid-derived bile acid. These increases correlated with faecal gymnemagenin levels.37

Other Activity
      Gymnema ingestion produced a significant lowering of cholesterol in a hypertension model, but did not lower (and even tended to increase) the raised systolic blood pressure induced by sugar feeding.38

      No information available.

      In acute toxicity studies no gross behavioural, neurologic or autonomic effects were observed in mice orally administered graded doses (250-8000 mg/kg) of an aqueous alcoholic concentrate of Gymnema (19.5:1). The LD50 value for this concentrate after oral administration was measured at 3.99 g/kg (equivalent to 78†g/kg of dried Gymnema leaf).33 To put this in context the LD50 for sugar in rats orally is 29.7 g/kg and for table salt is 3 g/kg.39 A substance is considered non-toxic if it produces no effect in a dose of up to 10 g/kg.40

Clinical Studies

      The reduction of urinary glucose levels in diabetics by oral administration of Gymnema was reported as early as 1926.41 The hypoglycaemic effect of Gymnema powder (10 g/day for 7 days) was investigated in 16 normal subjects and 43 mild diabetics in an uncontrolled trial. A hypoglycaemic effect was observed in the diabetics. Serum triglycerides, free fatty acids and cholesterol levels were also decreased. Excretion of creatinine was decreased in the diabetic group.42

      A controlled study on insulin-dependent diabetics found that a water-soluble Gymnema extract (400 mg/day corresponding to about 8 g of starting dried herb) reduced insulin requirements (by about 50%). Over the duration of treatment Gymnema lowered fasting mean blood glucose (by about 35%), glycosylated haemoglobin and glycosylated plasma protein levels from baseline values. Cholesterol was significantly reduced and brought to near normal levels. Triglycerides, free fatty acids and serum amylase were also lowered. The treatment period ranged from 6-30 months. The significant decrease in glycosylated haemoglobin occurred after 6-8 months of Gymnema treatment but remained significantly higher than normal values. None of these reductions were observed in control patients on insulin therapy alone who were studied over a period of 10-12 months. The authors suggested that Gymnema enhanced endogenous insulin production, possibly by pancreatic regeneration, as levels of C-peptide, a by-product of the conversion of proinsulin to insulin, were apparently raised (in comparison to both the insulin alone group and normal subjects).24

      A second study by the same research group found that the same Gymnema preparation (400†mg/day) produced similar results for non-insulin-dependent diabetics. Fasting blood glucose, glycosylated haemoglobin and glycosylated plasma protein were significantly reduced compared to baseline values (p<0.001) after 18-20 months of treatment. None of these reductions were observed in patients receiving conventional therapy alone who were studied over a period of 10-12 months. By the end of the treatment period cholesterol, triglycerides, phospholipids and free fatty acid levels were also significantly reduced compared to baseline values in those receiving Gymnema (p<0.001). Control patients receiving only conventional therapy achieved reductions in cholesterol, triglycerides and free fatty acids (p<0.05-p<0.001). Fasting and post-prandial serum insulin levels were significantly increased in the Gymnema group compared to those taking only conventional drugs (p<0.01). Twenty-one of the 22 patients were able to reduce their intake of hypoglycaemic drugs; 5 of these discontinued hypoglycaemic drugs entirely and maintained their blood glucose homoeostasis with Gymnema extract alone. The authors' suggestion of beta cell regeneration or repair facilitated by Gymnema was supported by the higher insulin levels in the serum of patients after Gymnema supplementation. Gymnema administration to healthy volunteers did not produce any acute reduction in fasting blood glucose level.25

      A clinical trial recently conducted in the US provides further support for the use of Gymnema in the management of diabetes. Of 65 patients tested over the 90-day trial, Gymnema tablets reduced mean fasting glucose levels by 11%. Average post-meal glucose levels showed a decline of 13% and glycosylated haemoglobin levels dropped 6.8%. In a subset of patients with the poorest control, results were more substantial. Pre-meal readings averaged an 18% decline, with post-meal levels reduced by 28%. Corresponding glycosylated haemoglobin levels declined 10%. Improved glucose control with Gymnema enabled 16% of the participants to decrease their prescription medication usage.43 The tablets used in the trial contained 400 mg of Gymnema extract (equivalent to around 4 g of leaf) standardised to 25% gymnemic acids. The dose used was two tablets per day.

Weight Loss
      A double-blind clinical trial investigated the effects of sweetness perception on short-term intake of food in men and women of normal weight.21 Participants were required to rate the 3 test solutions for sweetness (milkshakes with added sucrose, added aspartame or no added sweetener (placebo)). After an initial training in the sweetness scale, participants rinsed with either concentrated Gymnema extract solution (gymnemic acid content not defined) or placebo (tea solution) and rated the test solutions for sweetness. Subsequent ratings were made up to 60 minutes after the rinsing. Participants who had rinsed with Gymnema rated test solutions as less sweet compared to those who rinsed with placebo.

      Thirty minutes after the last set of ratings a test meal consisting of snacks was presented in the context of providing refreshment. Participants were told that this was not part of the experiment. The Gymnema group ate less total calories (501 ± 237 vs. 638 ± 333, p<0.006), total carbohydrates (p<0.003), total protein (p<0.018) and total fat (p<0.015) than participants whose taste perception was normal (tea group).


      The substantial assistance of Michelle Morgan in preparing this monograph is gratefully acknowledged.


1. Thakur RS et al. Major Medicinal Plants of India. Central Institute of Medicinal and Aromatic Plants, Lucknow, 1989.

2. Chopra RN et al. Chopra's Indigenous Drugs of India, 2nd Edn, 1958, reprinted Academic Publishers, Calcutta, 1982.

3. Kapoor LD. CRC Handbook of Ayurvedic Medicinal Plants. CRC Press, Boca Raton, 1990.

4. Shanmugasundaram KR et al. Pharmacol Res Commun 1981; 13: 475

5. Waller GR et al (eds). Saponins used in Food and Agriculture. Advances in Experimental Medicine and Biology, Volume 405. Plenum Press, New York, 1996.

6. Yoshikawa K. Tetrahedron Lett 1991; 32(6): 789-792

7. Yokota T et al. Nippon Shokuhin Kogyo Gakkaishi 1994; 41: 202

8. Yoshikawa K. Phytochem 1992; 31: 237

9. Kamei K et al. J Biochem (Tokyo) 1992; 111: 109

10. Sahu NP et al. Phytochem 1996; 41: 1181

11. Liu HM et al. Chem Pharm Bull (Tokyo) 1992; 40: 1366

12. Imoto T et al. Comp Biochem Physiol 1991; 100: 309

13. Warren RP et al. Nature 1969; 223: 94

14. Miyasaka A et al. Brain Res 1995; 676: 63

15. Hellekant G et al. Physiol Behav 1998; 65: 191

16. Min BC et al. Appl Human Sci 1998; 17: 9

17. Hellekant G et al. Physiol Behav 1996; 60: 469

18. Gent JF et al. Chem Senses 1999; 24: 393

19. Meiselman HL et al. Physiol Behav 1970; 5: 945

20. Frank RA et al. Chem Senses 1992; 17: 461

21. Brala PM et al. Physiol Behav 1983; 30: 1

22. Shanmugasundaram KR et al. J Ethnopharmacol 1983; 7: 205

23. Shanmugasundaram ER et al. J Ethnopharmacol 1990; 30: 265

24. Shanmugasundaram ER et al. J Ethnopharmacol 1990; 30: 281

25. Baskaran K et al. J Ethnopharmacol 1990; 30: 295

26. Kyremateng PH cited in Prendergast HDV (ed). Plants for Food and Medicine. Royal Botanic Gardens, Kew, 1998.

27. Yoshikawa M et al. Chem Pharm Bull 1997; 45: 2034

28. Shimizu K et al. J Smooth Muscle Res 1996; 32: 219

29. Wang LF et al. Can J Physiol Pharmacol 1998; 76: 1017

30. Shimizu K et al. J Vet Med Sci 1997; 59: 245

31. Okabayashi Y et al. Diabetes Res Clin Pract 1990; 9: 143

32. Srivastava Y et al. Int J Crude Drug Res 1986; 24: 171

33. Chattopadhyay RR. J Ethnopharmacol 1999; 67: 367

34. Sugihara Y et al. J Asian Nat Prod Res 2000; 2: 321

35. Yoshikawa M et al. Chem Pharm Bull 1997; 45: 1671

36. Teresawa H et al. Yonago Acta Med 1994; 37: 117

37. Nakamura Y et al. J Nutr 1999; 129: 1214

38. Preuss HG et al. J Am Coll Nutr 1998; 17: 116

39. Merck Chemical Databases online.

40. Rowan A. Of Mice, Models, and Men: A Critical Evaluation of Animal Research, State University of New York Press, Albany, 1984.

41. Gharpurey KG. Indian Med Gaz 1926; 61: 155

42. Balasubramaniam KB et al. J Natl Sci Counc Sri Lanka 1992; 20: 81

43. Joffe DJ, Freed SH. Newsletter for Professionals in Diabetes Care. Oct 31, 2001, Issue 76.



Search our pre-2001 archives for further information. Older issues of the printed magazine are also indexed for your convenience.
1983-2001 indices
; 1999-Jan. 2003 indices
Once you find the magazines you'd like to order, please use our convenient form, e-mail, or call 360.385.6021 (PST).

© 1983-2002 Townsend Letter for Doctors & Patients
All rights reserved.
Web site by Sandy Hershelman Designs
June 14, 2003