brand logo

Am Fam Physician. 2000;62(5):1051-1060

This is Part I of a two-part article on alternative therapies. Part II, “Congestive Heart Failure and Hypercholesterolemia,” will appear in the next issue.

Natural supplements are widely used in the United States and, while claims of their therapeutic effects abound, medical research does not always support their effectiveness. St. John's wort acts as a weak selective serotonin reuptake inhibitor with fewer side effects. S-Adenosylmethionine (SAMe) has enough of an antidepressant effect to warrant further research. More human studies are needed before garlic, bitter melon, soy and fenugreek supplements can be recommended for the management of diabetes, although chromium may be a promising treatment in some cases. Alpha lipoic acid is used in the treatment of diabetic neuropathy. The effects of ma huang/guarana combinations in obesity have not been well studied. These combinations may have potentially serious side effects but may also offer some benefit. The combination of hydroxycitric acid and garcinia has proved no more effective than placebo.

Because the number of family physicians recommending nutritional supplements to treat a variety of diseases is increasing, an awareness of the medical and legislative issues surrounding these products is critical.

In 1994, a grassroots campaign led to congressional passage of the Dietary Supplement Health and Education Act. This legislation restricted the authority of the U.S. Food and Drug Administration (FDA) over any herb or supplement as long as the manufacturer makes no claim that the product is able to affect a “disease state.” Under this legislation, a supplement is presumed safe until the FDA receives “numerous reports” of adverse effects. (From 1993 to 1998, the FDA received 2,621 reports of adverse effects linked to supplements. These complaints included reports of 101 deaths.) Family physicians must be aware of this lack of scientific oversight and must be on the alert for any side effects, including effects that have not yet been documented.1

It is also important to note that under this legislation, no standardization of the manufacturing of herbs or supplements is required. Although the supplement industry is currently moving to police itself in this regard, physicians and patients must be informed that supplements often contain varying amounts of active ingredients.1

American Family Physician has previously published articles that focus on herbal products and their side effects, as well as herb/drug interactions.2,3 For the most part, this article will focus on products not previously discussed in this journal. In this first part of a two-part article, we will highlight depression, diabetes and obesity, three conditions commonly encountered in the family physician's office, and discuss the role that natural products can play in the treatment of these medical conditions. Tables 1,2 and 3 summarize information about the natural products discussed.

ProductOther namesEfficacyMechanism of actionFormulationDosage*Side effectsDrug interactions
St. John's wortHypericum perforatum, hardhay, amber, goatweed, Klamath weed, Tipton weedWorks as well as TCAs for mild or moderate depressionActs as weak SSRI, dopaminergic, MAOI and norepinephrine uptake inhibitorDried seeds, leaves and flowers; capsule300 mg 3 times dailyRare photo-sensitivity rashMay lower digoxin level 25%; may contribute to serotonin syndrome if used with SSRIs/TCAs/MAOIs/dopamine agonists; may lower blood levels of protease inhibitors; may interfere with drugs used in organ transplantation
SAMeS-adenosylmethionine; AdoMetPossibly efficacious; poorly studied; cost prohibitiveMethyldonor important in neurotransmitter synthesisSynthetic form of a natural metabolite of the amino acid methionine; tablet400 to 1,200 mg per dayNone recordedNone recorded


The effectiveness and side effects of selective serotonin reuptake inhibitors (SSRIs) and tricyclic antidepressants (TCAs) are well documented. Relatively new potential additions to the armamentarium against depression include St. John's wort and S-adenosylmethionine (SAMe).


St. John's wort (Hypericum perforatum) appears to act as a weak SSRI and may have an affinity for sigma receptors as well. Its clinical effectiveness may be attributable to a combination of mechanisms because each mechanism by itself is too weak to provide the overall effect.4,5

Studies have compared TCAs (i.e., imipramine [Tofranil], amitriptyline [Elavil]) with St. John's wort in cases of mild to moderate depression and found no statistical differences between groups as evaluated by the Hamilton Depression Scale. One-third fewer side effects were noted in the subjects taking St. John's wort, compared with the subjects taking amitriptyline.6

Initially, it was thought that this botanical supplement rarely, if ever, led to any drug interactions.7 However, recent evidence8,9 indicates interactions with digoxin, theophylline (decreased bioavailability), protease inhibitors (lowering the level of indinavir [Crixivan]), and drugs used to combat transplant rejection (increasing the metabolism of cyclosporine [Sandimmune]) via induction of cytochrome P450 enzymes. It also may cause seratonin syndrome if used with other antidepressants.

Results of 29 clinical trials10,11 have demonstrated that St. John's wort is more effective than placebo in the treatment of mild to moderate depression. As yet, there is insufficient evidence comparing the efficacy of St. John's wort with conventional, modern antidepressants. We could find no studies that used St. John's wort in combination with other modalities to treat severe depression or treatment-resistant depression, and it should not be used in these patients.


Another potential antidepressant is SAMe, which is a molecule found naturally in all human cells. It plays a role in the many methylation reactions of the body, including gene expression, cell membrane homeostasis, and hormone and neurotransmitter synthesis.

The rationale for the use of SAMe in depression stems from its role in the metabolism of serotonin, dopamine and melatonin. Oral and intravenous SAMe supplementation has been shown to significantly increase SAMe levels in cerebrospinal fluid, indicating SAMe's crossover through the blood-brain barrier. This has been associated with increased levels of serotonin metabolites in cerebrospinal fluid. Some depressed patients may have low serotonin levels associated with low levels of SAMe.12,13

A 1994 meta-analysis14 performed in Italy found SAMe more efficacious than placebo in the treatment of depression and equally as effective as TCAs, with a much lower incidence of side effects. In a double-blind trial conducted that same year,15 26 patients received SAMe or desipramine for four weeks. At the end of the study period, 62 percent of the patients taking SAMe had significant improvement in symptoms, while only 50 percent of the patients taking desipramine had significant improvement. The study did not extend beyond this short observational period, leaving unanswered questions about the long-term side effects of SAMe and whether the antidepressant effects of SAMe persist beyond four weeks. The study did not take into account the usual four- to eight-week time period needed for TCAs to become maximally effective. The investigators did note, however, that all patients who showed a 50 percent decrease in their Hamilton Depression Scale scoring also showed a significant increase in plasma SAMe concentration.15

In 1995, an open multicenter study16 of 163 patients given 400 mg of SAMe per day showed a reduction in depressive symptoms at seven and 15 days. No serious side effects were observed. In their conclusion, the authors recommended that further double-blind trials be undertaken. Again, no documentation of long-term effectiveness or side effects has been reported. Finally, another review17 concluded that SAMe seemed to have enough of an anti-depressant effect to warrant further research.

In conclusion, because no studies have yet validated the long-term safety or efficacy of SAMe, further large long-term studies should be conducted before it receives widespread recommendation.

Diabetes Mellitus

The search for a cure for diabetes mellitus continues along traditional and alternative medicine fronts. Many herbal supplements have been used for the treatment of diabetes, but not all of them have scientific evidence to support their effectiveness. Bitter melon (Momordica charantia), fenugreek and soybeans have been studied as possible treatments in patients with diabetes, but the results of these studies were inconclusive or showed these products to be ineffective.1820 A summary of natural products used in the treatment of diabetes is given in Table 2.

ProductOther namesEfficacyMechanism of actionFormulationDosage*Side effectsDrug interactions
ChromiumNoneSomewhat efficacious; possibly needs more studyCounteracts insulin resistance; increases insulin receptor sensitivityPicolinate (organic form); capsule/tablet50 to 1,000 μg per dayIndigestion; case reports of renal failure with high dosesNone recorded
GarlicAllium sativum, poor man's treacleNone to date; few human studiesMay increase secretion of insulinFresh, oil, aqueous, fermented or driedVariesNone recordedNone recorded
Alpha lipoic acidThioctic acidEfficacious in decreasing symptoms of diabetic neuropathyAntioxidantCapsule600 to 1,200 mg per dayNone recordedNone recorded


Chromium picolinate is a commonly used nutritional supplement marketed in supra-nutritional dosages for the treatment of diabetes mellitus. It was discovered to have a hypoglycemic effect when added to the total parenteral nutrition given to diabetic patients in intensive care units.21 Subsequent animal studies revealed similar antidiabetic effects. There are also case reports in the medical literature that support its use in patients with type 1 diabetes mellitus (formerly known as insulin-dependent diabetes mellitus).22 The rationale for the use of chromium in type 2 diabetes mellitus (formerly known as non–insulin-dependent diabetes mellitus) stems from the fact that patients with type 2 diabetes lose more chromium in the urine than nondiabetic individuals; this phenomenon is thought to contribute to insulin resistance.23 A study involving 17 patients demonstrated that diets low in chromium may have negative effects on glucose tolerance in patients with borderline diabetes.24

While some studies have not demonstrated that chromium has any positive effect on serum glucose levels in humans,25 other placebo-controlled studies performed in China show beneficial effects when used in patients with type 2 diabetes.26 In one U.S. study of 180 patients with type 2 diabetes,27 subjects were subdivided into three groups: a placebo group and two groups taking different dosages of chromium supplements (i.e., 100 μg twice daily or 500 μg twice daily). Patients continued their usual medications and diet. At two months, the group receiving 500 μg of chromium picolinate twice a day experienced significant improvements in glycosylated hemoglobin (HbA1c) compared with the placebo group. After four months, subjects in both treatment groups had improved glucose levels, compared with the placebo group.27

Although initial human studies show positive effects, more studies are needed before chromium can be recommended as a supplement to improve serum glucose levels in patients with diabetes.


Many studies involving garlic have shown improved diabetes control in animals.28 S-allylcysteine sulfoxide, an active ingredient of garlic, was administered to alloxan-diabetic rats for a month, and their conditions improved compared with rats treated with glyburide and insulin.29 In another study, treatment of alloxan-diabetic rats with anti-oxidant S-allylcysteine sulfoxide isolated from garlic improved the diabetic condition to the same extent as glyburide and insulin.30 One possible mechanism of action of garlic on serum glucose may be direct: it may increase the pancreatic secretion of insulin from beta cells, or it may enhance insulin's release from its bound state.31 Other authors32 have suggested an associated insulin-sparing action. The only study33 identified using humans was a double-blind, randomized trial of 20 patients with type 2 diabetes performed in Thailand. The study did not show glucose-lowering effects of garlic compared with placebo. More human studies are needed before garlic supplements can be recommended for diabetic patients.


This fat-soluble, vitamin-like antioxidant is approved in Germany for treatment of diabetic neuropathy. It is found in liver, potatoes, broccoli and skeletal muscle. Alpha lipoic acid has no effect on levels of fasting serum glucose or HbA1c.

Animal studies have shown alpha lipoic acid to prevent nerve dysfunction, increase blood flow to nerves, help neural regeneration and improve distal nerve conduction rates.34

Subsequent human studies were conducted with the following results. In one study,35 328 patients with type 2 diabetes and symptomatic diabetic neuropathy were randomly assigned to treatment with intravenous doses of alpha lipoic acid: three doses of 1,200 mg, 600 mg or 100 mg of alpha lipoic acid per day or placebo taken over a three-week period. Symptoms measured included pain, burning, paresthesia and numbness in the feet. The total symptom score of the subjects decreased remarkably in the groups taking 600 mg and 1,200 mg, compared with the placebo group.

Another randomized, double-blind, placebo-controlled multicenter trial36 was designed to evaluate the efficacy and safety of oral alpha lipoic acid on cardiac autonomic neuropathy in patients with type 2 diabetes. The four-month study included 39 patients who took 800 mg per day of alpha lipoic acid and 34 patients who received placebo. Results indicated that alpha lipoic acid may slightly improve cardiac autonomic neuropathy in diabetic patients.

Hence, the current literature supports the use of alpha lipoic acid in diabetic neuropathy, but more studies are needed.


There are many antiobesity products on the market. Some of these products contain ma huang and guarana. A summary of natural products used in the management of obesity is given in Table 3.

ProductOther namesEfficacyMechanism of actionFormulationDosage*Side effectsDrug interactions
Ma huangEphedra, desert herb, ephedrinePossibly effective when used with guarana, but poorly studiedStimulant; naturally occurring ephedrineDried branchlets; extract; tablet20 mg ma huang/200 mg caffeine equivalent 3 times dailySame as ephedrine (e.g., tremor, agitation, insomnia, usually transient); linked to 17 cardiac deathsSame as ephedrine; additive effects with other stimulants
GuaranaPaullinia cupana, Brazilian cocoa, guarana breadPossibly effective when used with ma huang, but poorly studiedStimulant; naturally occurring caffeineSeeds ground and dried or roasted; paste to stick form; capsuleGuarana dosed as 200 mg caffeine equivalent/20 mg ma huang 3 times dailySame as caffeineSame as caffeine; additive effects with other stimulants
Ephedrine/caffeineEffective in studies up to 1 yearDecreased appetite; increased metabolic rateVaries20 mg/200 mg 3 times dailyTransient (8 weeks) mild tremor, agitation, insomniaAdditive effects with other stimulants
GarciniaGarcinia hanburyi, Camboge, Gutta Cambodia, Tom Rong, GambodiaNot efficaciousSaid to prevent body from storing fatResin dried to powderNone recordedStrong laxative effect, possible abdominal pain and vomiting at > 0.2 gNone recorded
Guggul gumGuggal resin, didin, didthin, gugalipidNot studiedSaid to release endogenous thyroid hormoneExtract; powdered resin; concentrated tabletsNone recordedNone reportedNone recorded


Ma huang contains ephedra, a compound that is similar to ephedrine and has similar effects. The chemical structure of the natural form of ephedrine is levorotatory, while the synthetic ephedrine is a racemic mixture. Ephedra has been shown to increase heart rate, affect blood pressure variably and increase 24-hour energy expenditure in humans.37 These effects are probably associated with the thermogenic effect of direct 38 beta1 and beta2 agonism.


Guarana is used by Brazilian Indians in a stimulating beverage similar to coffee or tea. It contains an exceptionally high level of caffeine, ranging from 3 to 5 percent by dry weight. In comparison, coffee beans contain approximately 1 to 2 percent caffeine, and dried tea leaves contain 1 to 4 percent caffeine.


No published studies have evaluated the potential for weight loss with guarana or ma huang. However, two unpublished studies have evaluated the effect of a ma huang/guarana combination (Metabolife).39 The first study showed a slight (3 to 6 percent) increase in metabolic rate over a 24-hour period but did not address weight loss. The second study, an eight-week trial of 60 patients, revealed that the combination promoted greater weight loss than placebo but stated that the study was too short to conclude anything about long-term efficacy or safety.39

In contrast to ma huang and guarana, much study has been devoted to caffeine and ephedrine. These two products used alone show no benefit in weight loss but appear to be beneficial when used in combination, although not without side effects. Ephedrine/caffeine combinations have been shown to reduce food intake, increase energy expenditure and decrease weight and body fat in animals.40 These findings have been corroborated in human open trials and human, randomized, double-blind, placebo-controlled trials.41 A randomized, double-blind, placebo-controlled study42 of 180 obese patients showed significantly greater weight loss with combination ephedrine/caffeine (20 mg/200 mg) over a 24-week period than with placebo.

One study found that 54 percent of patients treated with 20 mg/200 mg of ephedrine/caffeine experienced central nervous system side effects, especially agitation, but noted that these side effects declined markedly after the first month of treatment.43 In another study, the hemodynamic side effects of an ephedrine/caffeine combination (increased systolic but not diastolic blood pressure, increased heart rate) were transient while the thermogenic effects on energy expenditure were persistent.44

Finally, it should be remembered that ephedra has been linked to numerous episodes of toxicity, including 17 deaths.45 In addition to these published reports attributing sudden death to ma huang, there has also been a case report of hepatotoxicity, presumably associated with an unidentified contaminant in a ma huang product, and a case report of kidney stones induced by ma huang.46,47 No clinically relevant withdrawal symptoms were observed when subjects were discontinued from 20 mg/200 mg ephedrine/caffeine.41

In conclusion, although the weight loss effects of ma huang/guarana combinations have not been completely studied, the public continues to buy these products at an astonishing rate. We must be aware of the potential for beneficial effects as well as adverse effects of these diet aids.


Among other weight loss alternatives, hydroxycitric acid is often sold as the natural product garcinia. It is found in at least 14 over-the-counter weight loss products and has been shown to inhibit the enzymes that convert citrate into coenzyme A, essentially preventing the body from storing excess energy as fat. It has also been shown to inhibit appetite and reduce food intake in rats.48

One well-designed, double-blind, placebo-controlled study of the use of garcinia in humans has been published.49 This study found that garcinia was no more effective than placebo at the usual commercially available dose. The study has been faulted by some for measuring neither the appetite suppressant effect nor the bioavailability of hydroxycitric acid to the bloodstream. Opponents have postulated that the high-fiber diet used in the study may have limited the bioavailability, thus rendering the study ineffective and leading to the disappointing results.48 To date, however, the best available data indicate that this product is no more effective than placebo.


This resin from the myrrh species, in addition to being used as a cholesterol-lowering agent, is found in some over-the-counter diet products. It has been shown to stimulate the release of endogenous thyroid hormone in rats.50 We have found no studies supporting this claim in human or additional animal studies.

A list of resources for information about alternative therapies is given in Table 4. Resources for Information About Alternative Therapies

Some alternative/complementary medicine Web sites* that may be useful to family physicians are:
Alternative Medicine Foundation
The NCCAM Complementary and Alternative Medicine (CAM) Citation Index (CI)
American Botanical Council
Complementary and Alternative Medicine Program at Stanford (CAMPS)
Center for Complementary and Alternative Medicine Research in Asthma, located at the University of California, Davis
An extended list of relevant Web sites is also available in the October 1999 newsletter The Review of Natural Products (see below).
In addition, the following publications are recommended:
La Puma J, ed. Alternative Medicine Alert. Chicago: American Health Consultants, Inc., 2000.
Prescriber's Letter. Stockton, Calif.: Therapeutic Research Center, 2000.
The Review of Natural Products. St. Louis, Mo.: Facts and Comparisons, 2000.
(Telephone: 800-223-0554)
PDR for Herbal Medicines. 1st ed. Montvale, N.J.: Medical Economics, 1998.

*—These Web sites were verified August 2, 2000.

Continue Reading

More in AFP

Copyright © 2000 by the American Academy of Family Physicians.

This content is owned by the AAFP. A person viewing it online may make one printout of the material and may use that printout only for his or her personal, non-commercial reference. This material may not otherwise be downloaded, copied, printed, stored, transmitted or reproduced in any medium, whether now known or later invented, except as authorized in writing by the AAFP.  See permissions for copyright questions and/or permission requests.