Cholesterol Lowering Herbs Articles & Resources

* METHODS

Identification of clinical trials

To identify clinical trials involving herbal medicinal products with hypocholesterolemic properties, we conducted systematic literature searches in the following electronic databases (all from their inception to May 2001): MEDLINE (via PubMed), EMBASE, CINAHL, AMED (Alternative and Allied Medicine Database, British Library Medical Information Centre), the Cochrane Library (Issue 2, 2001), and CISCOM (Research Council for Complementary Medicine, London, UK). The search strategy is summarized in Appendix A (available online at .com).

Further relevant papers were located by hand-searching the reference lists of all papers and departmental files. In addition, experts in the field and manufacturers were contacted to provide published and unpublished material.

Inclusion and exclusion criteria Only randomized clinical trials investigating serum cholesterol reduction of monopreparations of herbal medicinal products administered as supplements were included. These could be placebo-controlled or equivalent trials. All retrieved data including uncontrolled trials, case reports, and preclinical and observational studies were reviewed for safety data. No language restrictions were imposed.

Data extraction and quality assessment

All articles were read in full. Data relating to sample size, study design, intervention and control, treatment duration, primary outcome measures, and results were extracted by the first author and validated by the second. The methodological quality of each trial was assessed using the Jadad scoring system, (3) which ranges from 0 (poorest) to 5 (highest). A score of 3 or above indicates reasonable methodological quality.

* RESULTS

We identified 11 herbal medicinal products investigated for hypocholesterolemic properties in randomized clinical trials: guggul (Commiphora mukul), artichoke (Cynara scolymus), garlic (Allium sativum), fenugreek (Trigonella foenumgraecum), red yeast (Monascus purpureus) rice, Asian ginseng (Panax ginseng), yarrow (Achillea wilhelmsii), eggplant (Solanum melongena), holy basil (Ocimum sanctum), milk thistle (Silybum marianum), and arjun (Terminalia arjuna).

The efficacy and safety of garlic has been reviewed extensively elsewhere (4-6) and is therefore not discussed in this paper. Details of all identified studies are shown in Tables 1-5 (and Table W1, available online at .com). Guggul, fenugreek, red yeast rice, and artichoke have been studied most extensively; randomized clinical trials of these herbal medicinal products with a Jadad score of 3 or above are discussed in more depth in Appendix B (available online at .com). Table 6 summarizes the adverse events experienced by subjects within these clinical trials and potential herb-drug interactions identified from systematic reviews.

Guggul (Commiphora mukul)

Six randomized clinical trials of guggul, involving 388 patients with different diagnoses, were identified. (7-12) Five were conducted in India and 1 in the United States; 4 were placebo-controlled; and 1 compared guggul with 2 reference compounds. The results suggest reductions in total serum cholesterol from 10% to 27% compared with baseline levels (Tables 1 and W1).

High-density lipoprotein (HDL) cholesterol levels were measured in 3 of the studies. (7-9) A significant increase was seen after 8 weeks of treatment in 1 study (9); in the others, no significant differences were seen. (7,8) A statistically significant decrease in lipid peroxide levels was reported in i study, with no corresponding change in the placebo-treated group. (7)

Several mild adverse events were reported during these trials, including rash, nausea, vomiting, eructation, hiccup, headache, loose stools, restlessness, and apprehension, although information regarding adverse events experienced during placebo administration was not always provided. A potential drug interaction with propranolol and diltiazem was investigated in a randomized crossover trial of 17 healthy volunteers, in which guggul was found to significantly reduce the peak plasma concentration of both drags. (13)

Fenugreek (Trigonella foenum-graecum)

Fenugreek seeds. Five randomized clinical triMs were identified, involving 140 patients; all but 1 trial was conducted in India (14-17) Although the methodological quality of the studies was considered generally poor in 4 of the trials, statistically significant reductions in total serum cholesterol of between 15% and 33% compared with baseline were demonstrated (Table 2).

Fenugreek leaves. In a single-blind study of 20 healthy male volunteers, Abdel-Barry and colleagues found a nonsignificant decrease of 9% in total serum cholesterol after a single dose of an aqueous extract made from fenugreek leaves (40 mg/kg) compared with a reduction of % after dilute coffee extract (placebo) (Table 2). (18)

Within all the identified studies of fenugreek, patients reported mild gastrointestinal symptoms such as increased flatulence, nausea, fullness, and diarrhea during fenugreek treatment, but none was severe enough to necessitate withdrawal from the study. A 14% reduction in serum potassium was noted in healthy volunteers after a single dose of an aqueous extract of fenugreek leaves. (18)

Red yeast rice

Red yeast rice is produced by solid-state fermentation of washed and cooked rice using the fungus Monascus purpureus. It has been used in Asia as a food preservative and colorant and for its medicinal properties since the Tang Dynasty (AD 800). It is available in capsules that contain a pulverized powder of fermented rice and yeast.

Four randomized clinical trials of the lipid-lowering effects of red yeast rice conducted in patients (n=695) with hyperlipidemia were identified (Table 3). (19-22) In all studies, statistically significant reductions (16% to 31%) in total serum cholesterol compared with placebo or control or baseline were seen.

Adverse events experienced in clinical trials included stomachache, heartburn, dizziness, and flatulence. No changes in liver function tests were demonstrated. There was 1 case report of a 26-year-old man who used red yeast rice in preparing sausages and developed anaphylaxis due to immediate sensitivity to M purpureus. (23) Whether this is relevant to the oral administration of red yeast rice capsules is not clear.

Artichoke (Cynara scolymus)

The choleretic effect of the leaf extract of artichoke has been studied widely, but only 2 randomized clinical trials of its hypocholesterolemic effects, involving 187 patients, were identified (Table 4). (24,25) One trial (n=44 healthy volunteers), published in abstract form only, found no significant difference in lipid levels compared with placebo, although post hoc subanalyses revealed some reductions in total serum cholesterol in patients with baseline levels above mmol/L; these results should be interpreted with some caution. Reductions in total cholesterol of % and % were reported in the other, larger trial after artichoke and placebo treatments, respectively.

No adverse events were reported during either study. Three post-marketing surveillance studies were located: one included 417 patients and reported excellent tolerability in 95%; in the second (203 patients) no adverse reactions were reported; and the third (553 patients) described mild adverse events in % of patients (flatulence, hunger, and weakness). (26-28)

* DISCUSSION

Many different herbal medicinal products have been identified with potential lipid-lowering properties, but the evidence for each herb is limited. The largest amount of published literature exists for guggul, fenugreek, red yeast rice, and artichoke, with reductions in total serum cholesterol ranging from 10% to 33%.

Although HDL and low-density lipoprotein (LDL) cholesterol were not measured in all the studies, increases in HDL and decreases in LDL levels were seen with guggul, red yeast rice, and yarrow, and decreases in LDL levels were seen in studies of fenugreek, arjun, and artichoke.

Safety

Few adverse events or drug interactions were reported in clinical trials of any of the 11 herbs identified. Many are used extensively in traditional medicine and culinary practices around the world, which supports their relative safety.

However, the long-term safety for use as herbal medicinal products has not been established. Long-term exposure of large numbers of patients within a formal setting would be necessary to determine safety, although difficulties associated with all herbal medicinal products exist, such as the inability to identify active ingredients and the potential for adulteration and misbranded products. No direct or indirect evidence exists for herb-drug interactions for fenugreek, guggul, Asian ginseng, and milk thistle (Table 6).

Study limitations

Although differences in study design, methodological quality, statistical methods, and subject populations create problems with interpretation of these figures, they appear to compare favorably with studies of garlic; the most recent metaanalysis suggested an average effect size of 4% to 6%. (6) Studies of conventional therapeutic options for hypercholesterolemia (eg, statins) have demonstrated reductions of 20% to 30% in serum cholesterol. (29)

Several shortcomings of the review need to be addressed. First, although attempts were made to obtain data from unpublished trials by contacting authors and manufacturers, none were located. There is evidence to suggest that studies with significant positive results are more likely to be published, (30) and this may be more pronounced with unfamiliar herbal therapies.

Second, because much of this research has been conducted in India and China, our extensive search strategy may not have located all the published material.

Third, there were several weaknesses with the original trialS; of the 25 randomized clinical trials of herbal medicinal products for serum cholesterol reduction identified, only 12 scored 3 or more points on the Jadad scale. The most frequent methodological flaws were conduct of single-blind or open studies and incomplete reporting of methods of randomization, blinding, and subject withdrawals.

Conclusions

Evidence suggests that physicians do not ask their patients about complementary and alternative therapies and that patients are reticent to discuss these treatments with their physicians. (31-34) Surveys indicate widespread use of complementary and alternative therapies among patients undergoing cardiac surgery. (1,2) Although no equivalent surveys have been conducted for patients with hypercholesterolemia, in light of the relatively large number of herbal medicinal products with potential lipid-lowering properties available, it seems prudent for physicians to explore this area in their clinical decision-making process.

In conclusion, although 11 herbal medicinal products were identified with potential hypocholesterolemic activity, the evidence supporting individual plants is limited. In addition to lowering cholesterol, several of the herbs may exert beneficial effects in cardiovascular disease by elevating HDL levels and inhibiting lipid oxidation. The safety profiles of the products in question seems to be encouraging. Further research is therefore warranted to establish the therapeutic value of these herbs in the treatment of hypercholesterolemia.

TABLE 1
Guggul: Randomized clinical trials for serum cholesterol reduction

 Jadad Subjects Treatment
Trial score (n) duration (wk)

Singh (7) 5 61 24
Kuppurajan (11) 5 40 3
 40
 40
Szapary (8) 3 103 8
Verma (9) 1 40 16
Bordia (12) 1 20 4
 20
Malhotra (10) 1 44 6-34

Trial Diagnosis Product

Singh (7) Hypercholesterolemia Gugulipid
 Placebo
Kuppurajan (11) Obesity Gum guggulu
 Guggulu A
 Clofibrate
 Placebo
 Hypercholesterolemia Gum guggulu
 Guggulu A
 Clofibrate
 Placebo
 Hyperlipidemia Gum guggulu
 Guggulu A
 Clofibrate
 Placebo
Szapary (8) Healthy subjects Gugulipid
 (75 mg/d)
 (150 mg/d)
 Placebo
Verma (9) Hyperlipidemia Gugulipid
 Placebo
Bordia (12) Healthy volunteers Guggulu A
 Placebo
 Coronary artery Guggulu A
 disease
 Placebo
Malhotra (10) Hyperlipoproteinemia Guggulu A
 EPC
 Ciba

 Change in lipid levels
 from baseline (%)
Trial TC HDL LDL

Singh (7) + 
 + +
Kuppurajan (11) -- --
 -- --
 -- --
 -- --
 -23 -- --
 -51 -- --
 -55 -- --
 -35 -- --
 -20 -- --
 +14 -- --
 -19 -- --
 -21 -- --
Szapary (8)
 -- * -- * +
 -- * -- * +
 -- * -- * 
Verma (9) + 
 + + +
Bordia (12) -- --
 + -- --
 -- --
 + -- --
Malhotra (10) -27 -- --
 -24 -- --
 -46 -- --

* Data not provided in abstract.

--, data not provided; TC, total cholesterol; HDL, high-density
lipoprotein cholesterol; LDL, low-density lipoprotein cholesterol;
Ciba, Ciba-13437-Su; EPC, ethyl-p-chlorophenoxy-isobutyrate

TABLE 2
Fenugreek: Randomized clinical trials for serum cholesterol reduction

 Jadad Subjects Treatment
Trial score (n) duration (wk)

Fenugreek seeds
Singh (14) 4 92 12
Prasanna (15) 2 18 3
Sharma (16) 2 15 
Sharma (16) 2 5 3
Sharma (17) 1 10 

Fenugreek leaves
Abdel- 3 20 Single dose
Barry (18)

Trial Diagnosis Product

Fenugreek seeds
Singh (14) Hypercholesterolemia Fenugreek
 Triphala *
 Placebo
Prasanna (15) Hypercholesterolemia Fenugreek
 (50 g)
 Fenugreek
 (100 g)
 Placebo
Sharma (16) NIDDM Fenugreek
 Standard diet
Sharma (16) NIDDM Fenugreek
 Standard diet
Sharma (17) NIDDM Fenugreek
 Standard diet

Fenugreek leaves
Abdel- Healthy volunteers Fenugreek
Barry (18) Control

 Change in lipid levels
 from baseline (%)-
Trial TC HDL LDL

Fenugreek seeds
Singh (14) 
 + 
 + 
Prasanna (15) 
 
 + +
Sharma (16) -- --
 -- -- --
Sharma (16) -- --
 -- -- --
Sharma (17) "TC and LDL were
 significantly reduced
 after the fenugreek
 diet as compared to
 the control diet. HDL
 was not significantly
 affected."

Fenugreek leaves
Abdel- -- --
Barry (18)
 -- --

* Triphala contains Emblica officinalis, Terminala chebula, and
Terminalia beterica.

--, data not provided; TC, total cholesterol; HDL, high-density
lipoprotein cholesterol; LDL, low-density lipoprotein cholesterol;
NIDDM, non-insulin-dependent diabetes mellitus

TABLE 3
Red yeast rice: Randomized clinical trials for serum
cholesterol reduction

 Jadad Subjects Treatment
Trial score (n) duration (wk)

Keithley (19) 3 92 8
Heber (20) 3 83 12
Shen (21) 2 152 8
Wang (22) 1 446 8

Trial Diagnosis Product

Keithley (19) HIV-related Red yeast rice
 dyslipidemia
 Placebo
Heber (20) Healthy volunteers Red yeast rice
 Placebo
Shen (21) Hypercholesterolemia Red yeast rice
 Control
Wang (22) Hyperlipidemia Red yeast rice
 Jiaogulan

 Change in lipid levels
 from baseline (%)
Trial TC HDL LDL

Keithley (19) 0 
 0 
Heber (20) 0 
 0 
Shen (21) + 
 0 +
Wang (22) + 
 + 

TC, total cholesterol; HDL, high-density lipoprotein cholesterol;
LDL, low-density lipoprotein cholesterol

TABLE 4
Artichoke: Randomized clinical trials for serum cholesterol reduction

 Jadad Subjects Treatment
Trial score (n) duration (wk)

Petrowicz (25) 5 44 12
Englisch (24) 3 143 6

Trial Diagnosis Product

Petrowicz (25) Healthy volunteers Artichoke
 Placebo
Englisch (24) Hyperlipoproteinemia Artichoke
 Placebo

 Change in lipid levels
 from baseline (%)
Trial TC HDL LDL

Petrowicz (25) -- * -- * -- *
 -- -- --
Englisch (24) 
 + 

* No data were provided for between-group comparisons. Details of
exploratory subanalysis are provided in the text.

--, data not provided; TC, total cholesterol; HDL, high-density
lipoprotein cholesterol; LDL, low-density lipoprotein cholesterol

TABLE 5
Other herbs: Randomized clinical trials for serum cholesterol
reduction

 Jadad Subjects Treatment
Trial score (n) duration (wk)

Guimaraes (35) 3 36 5
Guimaraes (35) 3 16 5
Gupta (38) 3 105 4
Sotaniemi (36) 2 36 8
Agarwal (39) 2 40 8
Asgary (40) 2 60 24
Petronelli (37) 1 20 8

Trial Diagnosis Product

Guimaraes (35) Hypercholesterolemia Eggplant
 Placebo
Guimaraes (35) Hypercholesterolemia Eggplant
 Placebo
Gupta (38) Coronary artery Arjun
 disease
 Vitamin E
 Placebo
Sotaniemi (36) NIDDM Asian
 ginseng
 (100 mg)
 Asian
 ginseng
 (200 mg)
 Placebo
Agarwal (39) NIDDM Holy basil
 Placebo
Asgary (40) Hypercholesterolemia Yarrow
 Placebo
Petronelli (37) Dyslipidemia Silymarin
 Placebo

 Change in lipid levels
 from baseline (%)
Trial TC HDL LDL

Guimaraes (35) + 
 
Guimaraes (35) + + +
 + 
Gupta (38) 
 + 
 + + +
Sotaniemi (36) NC NC NC
 NC NC NC
 -- NC NC
Agarwal (39) -- --
 + -- --
Asgary (40) -39 + 
 + +
Petronelli (37) + -- --
 -- --

--, data not provided; TC, total cholesterol; HDL, high-density
lipoprotein cholesterol; LDL, low-density lipoprotein cholesterol;
NC, no change; NIDDM, non-insulin-dependent diabetes mellitus

TABLE 6
Adverse events and possible herb-drug interactions (41-44)

 Possible interactions
Herb Adverse events (direct evidence) *

Guggul Rash, nausea, vomiting, Decreases peak plasma
 hiccup, headache, loose concentration of
 stools, restlessness, propranolol and diltiazem
 apprehension

Fenugreek Flatulence, diarrhea No direct evidence for
 interactions

Red yeast Stomach ache, head- No direct evidence for
rice ache, gastric discomfort interactions

Artichoke Flatulence, hunger, No direct evidence for
 weakness interactions

Eggplant None reported No direct evidence for
 interactions

Asian None reported Interaction with phenelzine
ginseng resulting in headache,
 tremor, and mania;
 interaction with warfarin
 decreases international
 normalized ratio;
 interaction with alcohol
 results in increased
 alcohol clearance

Milk thistle None reported Protection from
 hepatotoxins

Arjun None reported No direct evidence for
 interactions

Holy basil None reported No direct evidence for
 interactions

Yarrow None reported No direct evidence for
 interactions

 Possible interactions
Herb (indirect evidence) ([dagger])

Guggul Interference with drugs used
 for hypo- or hyperthyroidism,
 can increase effects of
 insulin and sulfonylureas

Fenugreek Increases effects of beta-
 blockers, calcium channel
 blockers, and cardiac
 glycosides; possible
 hypoglycemic activity

Red yeast No indirect evidence for
rice interactions

Artichoke No indirect evidence for
 interactions

Eggplant No indirect evidence for
 interactions

Asian Interaction with monoamine
ginseng oxidase inhibitors and with
 stimulants; increases effects
 of hypoglycemics

Milk thistle No indirect evidence for
 interactions

Arjun No indirect evidence for
 interactions

Holy basil No indirect evidence for
 interactions

Yarrow No indirect evidence for
 interactions

* Direct evidence: suspected interactions have been reported in case
reports and systematic clinical investigations.

([dagger]) Indirect evidence: interactions that are theoretically
possible and may have been seen in animal experiments but have
not been reported in a clinical situation.

ACKNOWLEDGMENTS

The authors thank Jongbae Park, Barbara Wider, and Francesca Borelli, Complementary Medicine, Peninsula Medical School, Universities of Exeter and Plymouth, for translation of papers from Chinese, Italian, and French, to and Esther Prati, Pharmaton, Lugano, for assistance with locating relevant articles. JTC received a research fellowship from Pharmaton SA, Lugano, Switzerland.

REFERENCES

(1.) Ai AL, Bolling SF. The use of complementary and alternative therapies among middle-aged and older cardiac patients. Am J Med Qual 2002; 17:21-27.

(2.) Liu EH, Turner LM, Lin SX, et al. Use of alternative medicine by patients undergoing cardiac surgery. J Thorac Cardiovasc Surg 2000; 120:335-341.

(3.) Jadad AR, Moore A, Carroll D, et al. Assessing the quality of reports of randomised clinical trials: is blinding necessary? Control Clin Trials 1996; 17:1-12.

(4.) Warshafsky S, Kamer RS, Sivak SL. Effect of garlic on total serum cholesterol. A meta-analysis. Arch Intern Med 1993; 119:599-605.

(5.) Silagy C, Nell A. Garlic as a lipid lowering agent: a meta-analysis. J R Coll Physicians Lond 1994; 28:39-45.

(6.) Stevinson C, Pittler MH, Ernst E. Garlic for treating hypercholesterolemia. A meta-analysis of randomised clinical trials. Ann Intern Med 2000; 133:420-429.

(7.) Singh RB, Niaz MA, Ghosh S. Hypolipidemic and antioxidant effects of Commiphora mukul as an adjunct to dietary therapy in patients with hypercholesterolemia. Cardiovasc Drags Ther 1994; 8:659-664.

(8.) Szapary PO, Wolfe ML, Bloedon LT, Fair MB, Berezich DJ, Cirigliano AJ, Rader DJ. A double blind, randomised, placebo controlled clinical trial of standardized guggul extract in patients with hypercholesterolemia. Complement Ther Med 2002;10:112.

(9.) Verma SK, Bordia A. Effect of Commiphora mukul (gum guggulu) in patients with hyperlipidemia with special reference to HDL cholesterol. Indian J Med Res 1988; 87:356-360.

(10.) Malhotra SC, Ahuja MMS. Comparative hypolipidaemic effectiveness of gum guggulu (Commiphora mukul) fraction 'A', ethyl-P-Chlorophenoxyisobutyrate and Ciba-13437-Su. Indian J Med Res 1971; 59:1621-1632.

(11.) Kuppurajan K, Rajagopalan SS, Koteswara Rao T, Sitaraman R. Effect of guggulu (Commiphora mukul-Engl) on serum, lipids in obese, hypercholesterolemic and hyperlipemic cases. J Assoc Physicians India 1978; 26:367-373.

(12.) Bordia A, Chuttani SK. Effect of gum guggulu on fibrinolysis and platelet adhesiveness in coronary heart disease. Indian J Med Res 1979; 70:992-996.

(13.) Dalvi SS, Nayak VK, Pohujain SM, Desai NK, Kshirsagar NA, Gupta KC. Effect of gugulipid on bioavailability of diltiazem and propranolol. J Assoc Physicians India 1994; 42:454-455.

(14.) Singh RB, Niaz MA, Rastogi V, Singh N, Postiglione A, Rastogi SS. Hypolipidemic and antioxidant effects of fenugreek seeds and triphala as adjuncts to dietary therapy in patients with mild to moderate hypereholesterolemia. Perfusion 1998; 11:124-130.

(15.) Prasanna M. Hypolipidemic effect of fenugreek: a clinical study. Indian J Pharmacol 2000; 32:34-36.

(16.) Sharma RD, Raghuram TC. Hypoglycaemic effect of fenugreek seeds in non-insulin dependent diabetic subjects. Nutr Res 1990; 10:731-739.

(17.) Sharma RD, Raghuram TC, Rao NS. Effect of fenugreek seeds on blood glucose and serum lipids in type I diabetes. Eur J Clin Nutr 1990; 44:301-306.

(18.) Abdel-Barry JA, Abdel-Hassan IA, Jawad AM, Al-Hakiem MHH. Hypoglycaemic effect of aqueous extract of the leaves of Trigonella foenum-graecam in healthy volunteers. East Mediterr Health J 2000;6:83-88.

(19.) Keithley J, Swanson B, Sha B, Zeller J, Kessler HA, Smith KY. A pilot study of the safety and efficacy of Cholestin in treating HIV-related dyslipidemia. Nutrition 2002; 18:201-204.

(20.) Heber D, Yip I, Ashley JM, Elashoff DA, Elashoff RM, Go VLW. Cholesterol lowering effects of a proprietary Chinese red yeast rice dietary supplement. Am J Clin Nutr 1999; 69:231-236.

(21.) Shen Z, Yu P, Su M, et al. A prospective study of Zkital capsules in the treatment of primary hyperlipidemia. Natl Med J China 1996; 76:156-157.

(22.) Wang J, Lu Z, Chi J, et al. Multicenter clinical trial of the serum lipid-lowering effects of a Monascus purpureus (red yeast) rice preparation from traditional chinese medicine. Curr Ther Res 1997; 58:964-978.

(23.) Wigger-Alberti W, Bauer A, Hipler UC, Elsner P. Anaphylaxis due to Monascus purpureus fermented rice (red yeast rice). Allergy 1999; 54:1328-1336.

(24.) Englisch W, Beckers C, Unkauf M, Ruepp M, Zinserling V. Efficacy of artichoke dry extract in patients with hyperlipoproteinemia. Arzneimittelforschung 2000; 50:260-265.

(25.) Petrowicz O, Gebhardt R, Donner M, Schwandt P, Kraft K. Effects of artichoke leaf extract (ALE) on lipoprotein metabolism in vitro and in vivo. Atherosclerosis 1997; 129:147.

(26.) Fintelmann V. Antidyspeptic and lipid lowering effect of an extract from artichoke leaves, results of clinical trials on efficacy and tulerability of Hepar-SL in 553 patients. Z Allg Med 1996; 72:3-19.

(27.) Fintelmann V, Wegener T. Langzeitanwendung von artischockenblattertrockenextrakt (Hepar-SL forte) bei dyspeptischem symptomkomplex. Presented at: Phytotherapie Kongress 1997; November 27-28, 1997; Wurzburg, Germany.

(28.) Held C. Vonder 1. Deutsch-ungarischen Phytopharmakakonferenz 1991; 20 November, Budapest. Z Klin Med 1992; 47:92.

(29.) Scandinavian Simvastatin Survival Study Group. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 1994; 344:1383-1389.

(30.) Easterbrook PJ, Berlin JA, Gopalan R, Matthews DR. Publication bias in clinical research. Lancet 1991; 337:867-872.

(31.) Adler SR, Fosket JR. Disclosing complementary and alternative medicine use in the medical encounter: a qualitative study in women with breast cancer. J Fam Pract 1999; 48:453.

(32.) Eisenberg DM, Davis RB, Ettner SL, et al. Trends in alternative medicine use in the United States, 1990-1997. Results of a follow-up national survey. JAMA 1998; 280:1569-1575.

(33.) Eisenberg DM, Kessler RC, Foster C, Norlock FE, Calkins DR, Delbanco TL. Unconventional medicine in the United States. N Engl J Med 1993; 328:246-252.

(34.) Abbot NC, Ernst E. Patients' opinions about complementary medicine. Forsch Komplementarmed 1997; 4:164-168.

(35.) Guimaraes PR, Galvao AMP, Batista CM, et al. Eggplant (Solanum melongena) infusion has a modest and transitory effect on hypercholesterolemic subjects. Braz J Med Biol Res 2000; 33:1027-1036.

(36.) Sotaniemi EA, Haapakoski E, Rautio A. Ginseng therapy in non-insulin dependent diabetic patients. Diabetes Care 1995; 18:1373-1375.

(37.) Petranelli A, Roda E, Briganti M, Labate AMM, Barbara L. Effeto della somministrazione di silimarina sui livelli dei lipidi sierici. Clin Ter 1981; 99:471-482.

(38.) Gupta R, Singhal S, Goyle A, Sharma VN. Antioxidant and hypocholesterolaemic effects of Terminalia arjuna tree bark powder: a randomised placebo controlled trial. J Assoc Physicians India 2001; 49:231-235.

(39.) Agarwal P. Rai V, Singh RB. Randomised placebo-controlled, single blind trial of holy basil leaves in patients with non insulin dependent diabetes mellitus. Int J Clin Pharmacol Ther 1996; 34:406-409.

(40.) Asgary S, Naderi GH, Sarrafzadegan N, Mohammadifard N, Mostafavi S, Vakili R. Antihypertensive and antihyperlipidemic effects of Achillea wilhelmsii. Drugs Exp Clin Res 2000; 26:89-93.

(41.) Fugh-Berman A. Herb-drug interactions. Lancet 2000; 355:134-138.

(42.) Ernst E. Possible interactions between synthetic and herbal medicinal products. Part 1: a systematic review of the indirect evidence. Perfusion 2000; 13:4-15.

(43.) Ernst E. Interactions between synthetic and herbal medicinal products. Part 2: a systematic review of the direct evidence. Perfusion 2000: 13:60-70.

(44.) Thompson Coon J, Ernst E. Panax ginseng: a systematic review of adverse effects and drug interactions. Drug Saf 2002; 25:323-44.

Corresponding author: Joanna Thompson Coon, PhD, BSc (Hons), Complementary Medicine, Peninsula Medical School, Universities of Exeter and Plymouth, 25 Victoria Park Road, Exeter EX2 4NT, UK. E-mail: .uk.

Joanna S. Thompson Coon, PhD, BSc (Hons) and Edzard Ernst, MD, PhD, FRCP (Edin)

Peninsula Medical School, Universities of Exeter and Plymouth, Exeter, United Kingdom