But he is imprudent in encouraging the use of ginkgo biloba, advocating large doses and setting no upper limit on the amount readers should take. He never warns about this herb’s anticoagulant, or "blood-thinning," properties, a matter of concern to people who might require emergency surgery or who are being treated with blood thinners.

Atkins is not alone in promoting dietary supplements. They are advertised everywhere, for all manner of ailments, and about one-third of Americans buy herbal products like echinacea, ginseng, and St. John’s wort. They spend about $5 billion annually at retail outlets, and sales are rising about 18 percent a year. Thousands of products cram the shelves of health food stores, grocery markets, and pharmacies nationwide. They are also widely available through catalogs and the Internet. Even major pharmaceutical companies are adding dietary supplements to their lines. But what assurances do consumers have about exactly what they are getting, and about the safety of the products?

Regulation of dietary supplements varies considerably among developed countries, even among member states of the European Union, which has yet to adopt any transnational standards for these products. Under the most extensive legal framework, Germany has tested more than 300 herbal remedies since 1980, finding about two-thirds of these products to be safe and at least minimally effective (under a very liberal standard). It controls these substances as drugs. But consumers in other Western societies receive little protection.

Last year, Canada allotted $7 million (Canadian) to establish an Office of Natural Health Products over a three-year period. Although this new agency will govern the premarket assessment, labeling, licensing, and monitoring of herbal supplements, the precise scope of its mandate has yet to be determined. In the meantime, these products are blooming and booming in a regulatory vacuum.

In Britain, the majority of herbal remedies are classified as food supplements and are thus unlicensed. Others, sold by herbalists, are specifically exempt from licensing under a 1968 law. Only a small number of herbs that are regulated as drugs by the Medicines Control Agency carry any real assurance that they are safe and effective. But few places in the industrialized world, if any, have a more permissive environment than the United States.

The U.S. Congress has virtually exempted herbal remedies from government oversight. When the Food and Drug Administration considered regulating these products in the early 1990s, manufacturers and health food stores orchestrated a massive lobbying campaign against stricter controls. The industry produced television commercials that depicted movie star Mel Gibson handcuffed by FDA agents for possessing vitamins. The result was the Dietary Supplement and Health Act of 1994. Pushed heavily by Sen. Orrin Hatch of Utah, the home base of many supplements makers, and passed over the objections of the FDA, the law created a new product class, the dietary supplement, which was not subject to regulations applied to drugs. Now any substance that can be found in foods, regardless of amount or action and including chemicals that act as hormones or toxins, can be produced and sold without any premarket testing or approval.

The FDA can restrict the sale of an herbal product only if the agency receives well-documented reports of health problems associated with it. The FDA formed its Special Nutritionals Adverse Event Monitoring System in 1993 as an important component of its MEDWatch program, which is designed to track problems arising from the range of drugs and medical devices the fda is charged with regulating. The Special Nutritionals monitoring system was established to record adverse reactions associated with the more than 3,000 dietary supplement products that were then on the market.

But sometime around October 1998, after logging 2,621 "adverse reactions," including 184 fatalities, associated with these products, the FDA quietly stopped monitoring most cases in which herbal remedies had been linked to illness or death. Top agency officials said that they cut the monitoring program because of budgetary constraints. Although the FDA insists that it still collects data on medical emergencies linked to dietary supplements, only three staff members are now assigned to evaluate adverse reactions caused by the industry’s products, which are made by hundreds of private companies.

The agency and its herbal monitoring system have been criticized by the congressional General Accounting Office. In July 1999, the GAO said that the FDA’s adverse-reaction reports were incomplete, poorly documented, and inadequate to use as a basis to establish safe dosages for herbal supplements. The FDA requested an additional $2.5 million specifically to improve its dietary supplements tracking program in its fiscal 2000 budget, but this was among several FDA requests eliminated by a Senate subcommittee.

The number of medical emergencies linked to dietary supplements has unquestionably been underreported. Unlike the monitoring system for regulated drugs, which requires manufacturers to report side effects, the dietary supplement system relies on voluntary reports of adverse reactions from the makers of herbal products. This is rather like the IRS asking taxpayers to voluntarily provide information on their own underreporting of income, and some manufacturers have failed to report thousands of medical problems linked to their brands. Depositions in a recent court case involving the death of a woman from Sacramento, Calif., show that E’ola, a Utah multilevel marketing firm, received 3,500 customer complaints about its Amp II, a diet formula containing the herbal stimulant ephedra. The San Francisco Chronicle reported that none of those incidents was ever disclosed to the FDA.

In a regulation that went into effect earlier this year, the FDA worsened the situation by freeing supplement manufacturers to make all manner of dubious health claims — that their products treat conditions such as premenstrual syndrome and acne, for example, although there is little or no evidence to support these assertions.

With FDA authority limited by the 1994 law, the Federal Trade Commission, which monitors advertising, has taken a more vigorous role in regulating the makers of supplements. In 1999, the FTC took legal action against seven manufacturers that had violated regulations requiring advertising to be truthful and verifiable. The companies were selling cure-alls for conditions such as impotence, cancer, and obesity. The commission also sent e-mail warnings to 1,200 Internet sites that it said had made "incredible claims" for drugs, devices, and supplements, including herbal remedies that would supposedly ward off aids. Late last year, the commission also issued its first set of advertising guidelines aimed specifically at the supplement industry.

Even with these FTC measures, however, consumers have no real guarantee that dietary supplements are safe or effective, that the information about dosage on the label is correct, or even that the substance in the container is genuine. Only a few herbal supplements — saw palmetto for treating enlarged prostate glands and ginkgo for improving memory slightly in Alzheimer’s patients — have been shown to be at all efficacious.

Moreover, because the demand for some herbal remedies exceeds the supply from natural sources, unscrupulous manufacturers occasionally have switched to cheaper look-alike substances that are pharmacologically different. In February 2000, researchers in Britain discovered that eight brands of herbal skin ointments sold as "natural" treatments for eczema illegally contained dexamethasone, a potent prescription steroid. The creams had such huge quantities of steroid that they could have permanently damaged delicate skin. The creams for children, whose skin is most vulnerable, contained five times more steroid than the adult products. At about the same time, the FDA ordered four California supplement distributors to recall herbal compounds that illegally contained glyburide or phenformin, prescription drugs that lower blood-sugar levels. The action came after a diabetic in California developed hypoglycemia after taking a product said by its manufacturers to contain only natural Chinese herbs.

When ConsumerLab.com, a new private company, tested various herbal remedies, it said that many of the products did not deliver what the manufacturers promised. It found fault with 10 of 27 brands of saw palmetto studied, and discovered that in one-quarter of the 30 brands of ginkgo biloba, the levels of the active ingredient were less than indicated on the labels. Chemists at the Good Housekeeping Institute recently analyzed eight brands of SAMe, an herbal preparation advertised as a "natural Prozac" to relieve depression, and found that two had only half the promised levels of active ingredient, while another contained none at all. Consumer Reports also examined 10 brands of ginseng and concluded that several contained almost none of the active ingredient.

There is no shortage of information available to consumers about dietary supplements, but it is heavy on advocacy and light on scientific proof. Putting it another way, until there is evidence that rises to the standards of the medical and scientific communities that a preventative or therapeutic nostrum actually works, we do not know that it does work. We might just as well take a pinch of herbes de Provence along with a multivitamin pill each morning.

Worse yet, many of these products are already known to be toxic, carcinogenic, or otherwise hazardous, causing high blood pressure, deadly allergic reactions, cardiac arrhythmias, and kidney or liver failure. Some also can exacerbate autoimmune diseases like arthritis and lupus. In February, articles in the medical journal the Lancet identified life-threatening side effects of St. John’s wort: interference with the protease inhibitor indinavir, which can lead to treatment failure in aids patients; and rejection of heart transplants by an interaction with the immune suppressant cyclosporine.

In June, the New England Journal of Medicine reported that a Chinese herb that caused kidney failure in dozens of Belgian dieters in the early 1990s appears to have even deadlier long-term effects — cancer and precancerous lesions. The subjects in this study were an unlucky subset of some 10,000 Belgian dieters, who from 1990 to 1992 took a combination of Chinese herbs and conventional drugs prescribed by weight-loss clinics. After dozens of victims suffered kidney failure, investigators discovered that Belgian pharmacists had been using mislabeled Chinese herbs to produce the diet pills. Instead of using Stephania tetrandra, the druggists had filled the pills with derivatives of the herb Aristolochia fangchi, which has long been known to damage kidneys and to cause cancer in animals. At least 70 people experienced complete kidney failure, while another 50 suffered kidney damage serious enough to require treatment.

The first urinary tract cancers were found among these victims in 1994. To prevent the onset of the disease in others, doctors in Brussels advised patients whose kidneys and ureters had stopped functioning to have the organs surgically removed. Thirty-nine people chose to have the operation over the past several years. When researchers studied the excised tissue, they were shocked to discover that 18 of the patients had already developed cancer, while 19 others had precancerous lesions.

Belgium banned the importation of Aristolochia in 1992. In May 2000, the FDA distributed warnings to health professionals and the supplements industry about the dangers of Aristolochia, and it plans to block the herb’s entry into the United States in the near future. For many critics of current U.S. policies on dietary supplements, this action is long overdue, considering that Germany banned Aristolochia in 1981 and that the World Health Organization issued a warning about the herb in 1982. Medical organizations also have advised people to stop using herbal remedies such as St. John’s wort, ginkgo biloba, and ginseng at least two or three weeks before any scheduled surgery to avoid dangerous interactions with the anesthesia.

To the growing alarm of medical experts, dietary supplements companies have begun aggressively marketing their products to children and their parents. They peddle a variety of potent concoctions that supposedly will help kids gain strength or lose weight, or will treat illnesses ranging from colds and flu to depression and even Attention Deficit Disorder (ADD). As a result, increasing numbers of children are gulping supplements, often at the insistence of parents in search of "natural" remedies or "healthful" alternatives for youngsters who eat too much junk food. In 1999, a study conducted by National Public Radio, the Kaiser Foundation, and the Kennedy School of Government found that 18 percent of parents were giving their children dietary supplements other than vitamins or minerals.

In Vancouver, Wash., Nutrition Now Inc., for example, uses a cute rhinoceros cartoon character to promote its line of dietary supplements for kids, including Rhino Pops, an echinacea-based cold remedy. From Saco, Maine, Fresh Samantha Inc. ships to grocery stores nationwide Body Zoomer fruit smoothies that carry cartoon pictures of children to catch the eye. Their drinks contain ginseng, guarana, and spirulina (a green algae).

Although many supplement companies advise youngsters under 18 not to take the stimulant ephedra, many products contain this toxic ingredient, and a few industry representatives still recommend it for kids with add. In a report earlier this year, the fda documented 134 cases linking ephedra to serious reactions, including insomnia, nervousness, seizure, hypertension, stroke, and death, over a 33-month period ending in March 1999. Ten of the reports involved children younger than 18. But the manufacturers’ position remains ambiguous. The industry opposes a proposed New York ordinance that would prohibit ephedra sales to minors, claiming that "it would be more effective as a labeling issue," said Wes Siegner, counsel to the ephedra committee of the American Herbal Products Association. "You don’t want to cause trouble for salesclerks." Despite the warning labels, however, ephedra still enjoys considerable popularity among teenagers trying to lose weight.

Supporters of herbal supplements who scoff at reports of adverse reactions retort that with tens of millions of people consuming them in this country alone, there would be a significant body count if the products were genuinely dangerous. There are several answers to such an assertion. Even with profound underreporting of problems, reports of serious adverse effects are in fact mounting. In the absence of compelling evidence of benefit from the use of the vast majority of herbal supplements, these severe — even life-threatening — side effects are particularly worrisome (some would say unacceptable). Finally, problems are likely to increase with some large retailers now selling new higher-potency formulations and consumers taking large doses for long periods of time.

The supplement industry is moving voluntarily toward standards for "good manufacturing practices," a concept borrowed from drug makers that includes "adequately equipped manufacturing facilities, adequately trained personnel, stringent control over the manufacturing processes, reliable and secure computerized operations, and appropriate finished product examination and testing." But these measures address only the identity, purity, and potency of the products, neglecting the fundamental question of whether the active ingredients themselves are safe and effective, and they offer no independent verification that the standards have been met.

Regulation of the herbal supplement industry is so lax, and some of the products so dangerous, that a public health catastrophe seems inevitable. For example, a mislabeled or contaminated product could cause numerous cases of organ failure or death. After such a major incident, Congress would almost certainly move quickly to reclassify herbal supplements as drugs, which are stringently regulated by the FDA — requiring a demonstration of safety and efficacy before they could be marketed. Every product would have to gain approval from the FDA, a process that takes 12 to 15 years and costs, on average, about $500 million. Few of the herbal remedies now in stores could meet such criteria, and consumers would thus be denied access to many of them.

It is tempting to suggest, as some have, that herbal supplements might be provided a regulatory shortcut, in the form of a requirement for certification only of safety, but this proposal presents difficulties. For one thing, the safety of a drug (or herbal supplement) in normal clinical use is difficult to judge in a vacuum; a global judgement of safety and efficacy together is often necessary. The level of risk — side effects, or adverse reactions — that is acceptable varies widely and depends on the product’s particular use, or "indication." For example, severe and frequent side effects that might be tolerable in a drug offering a definitive cure for AIDS or pancreatic cancer would be unacceptable if the product were intended to treat baldness or hay fever symptoms.

The current unregulated commerce in dietary supplements may pose a real threat to public health, and even those who believe that the new drug approval process is too stringent would not argue for the complete absence of oversight. Before a calamity occurs, the worldwide industry should find better ways to police itself.

That models for oversight exist? The FDA enjoys an absolute monopoly over regulation, a situation that can produce delays, increased expense, and inefficiency. A wide range of other institutional models for product oversight exists. At the opposite pole from a government monopoly is a laissez-faire, self-regulating system in which government has no role at all and decisions about testing and marketing are made unilaterally by manufacturers. This is the system that currently exists for dietary supplements. Between the two extremes, other institutional alternatives offer various configurations of government oversight and nongovernmental mechanisms that could regulate and monitor the nation’s supply of dietary supplements. These alternative organizational arrangements involve varying degrees of privatization and FDA control of the certification process.

Recent history provides evidence of the failures of the extremes of monopoly and laissez-faire to serve the public interest; abuse of power is too easy in both. Government monopolies can foster abuses by bureaucrats. Laissez-faire creates similar temptations for companies, with more obvious public risk. Proponents of a laissez-faire system maintain that in the absence of government oversight, companies would be induced to protect the integrity of their products by goodwill, a desire to preserve their reputations, and the fear of civil litigation or even criminal prosecution. But while an acceptable level of safety might be achieved much of the time, inevitably some manufacturers would take dangerous shortcuts in production, testing, or surveillance. Public reaction to isolated incidents of poisonings led Congress to pass the Biological Act of 1902 and the Food, Drug, and Cosmetic Act of 1938. This history illustrates some of the pitfalls of a laissez-faire system.

If the extremes are unworkable, what other models are available? One possible arrangement would be a private monopoly organization to review and certify new products, perhaps under the sanction of the FDA or some other federal agency. Still another would have supplement manufacturers contracting with competing private entities to perform the certification.

Companies could elect to contract voluntarily with a newly established foundation that could operate like Underwriter’s Laboratories (UL), a large, not-for-profit organization that tests and certifies more than 16,500 types of products. Many of these items, such as electrical appliances and equipment, automotive and mechanical parts, fire-resistant building materials, and bullet-resistant glass, present inherent potential hazards to life and property.

Using more than 650 discrete standards, or guidelines, certification from ul or its competitors currently offers assurance of safety but not of effectiveness, except in a few special cases where the two factors are inextricably linked, such as fire extinguishers and smoke detectors. These organizations charge a fee based on the cost of developing standards and testing, which in turn depends on the characteristics of the particular published standard for the product in question. UL does not underwrite any risk of the products it certifies, nor are its liability concerns excessive. As an independent third party, UL is not jointly liable for any defects in the product’s performance in situ, and UL certification does not connect the organization to any negligence by the manufacturer. UL and its rivals hold no monopoly on safety certification, and their endorsements are not equivalent to government approval. Many retailers, however, are reluctant to carry products that lack certification by UL or one of its competitors. Insurers also occasionally deny liability coverage for products without it. Companies, insurers, and the independent testing laboratories all have incentives to maintain high standards both for consumer products and for manufacturers.

Another similar model for voluntary self-regulation that functions routinely and efficiently in the United States is the certification of plant seeds sold to agricultural producers or growers in order to prevent any compromise of seed quality or consistency. In California, for example, oversight is performed by the nonprofit California Crop Improvement Association (CCIA), which provides a voluntary quality assurance program for the maintenance and increase of crop seed. Each variety that enters this program is evaluated for its unique characteristics such as pest resistance, adaptability, uniformity, quality, and yield. Seed production is closely monitored by CCIA to prevent out-crossing, weed, other crop, and disease contamination that may negatively affect seed quality. Seed movement is monitored from field harvest, through the conditioning plant, and into the bag. Samples can be rejected if "off-type" seeds are found at a percentage that is greater than standards permit, as is occasionally the case with beans, cereals, and sunflowers.

Another, perhaps even more apposite, model for regulating the supplement industry is the procedure for regulating the majority of medical devices in the European Union. In it, competing private entities perform certifications.

This system relies heavily on various sets of product standards and normally does not involve government regulators directly in product oversight. For low risk (so-called Class I) devices, such as tongue depressors and eyeglasses, manufacturers themselves are allowed to certify that their products meet the necessary standards. For higher risk, Class II devices, those that might inflict serious harm if they malfunction but that are unlikely to kill patients, manufacturers must obtain third-party review from private sector, profit-making "notified bodies," which test products, inspect manufacturing systems, and ultimately verify that EU standards have been met. Following this certification, the products can be marketed. Since 1995, most Class II devices, such as X-ray machines, dentists’ drills, heartbeat monitors, and hip implants, have undergone this sort of nongovernmental third-party examination within the EU.

Because these "notified bodies" vie with each other for business, they have reason to be expeditious but thorough in their inspections. As a result, approval of new medical devices in Europe takes perhaps half as long as it does in the United States, shortening the overall development process by roughly two years, apparently without compromising safety. The more favorable regulatory climate within the EU has allowed many companies to establish a positive cash flow by introducing new products in Europe before marketing them in the United States. The British consulting firm the Wilkerson Group has compiled a list of 100 medical devices that are already available to patients in other developed nations, but not in the United States. This trend is most evident in the fast-moving areas of medicine, such as imaging and cardiology. Recently, Medtronics, one of the largest U.S. device makers, announced that it was transferring its entire venture group to Europe.

Voluntary mechanisms of self-regulation such as UL and CCIA could be adapted by the manufacturers of dietary supplements for effective oversight of their products, without involving federal regulators. UL and similar organizations are sanctioned by the Occupational Safety and Health Administration (OSHA), an agency of the U.S. Department of Labor. Such a government mandate arguably would not be necessary, however, as long as manufacturers of dietary supplements maintained an arm’s length relationship from the regulator. In other words, the overseer should be a certifier, not a collaborator.

ConsumerLab.com, the new private testing firm mentioned earlier, is an example of the model according to which supplement manufacturers contract with competing private entities — or at least it would be if there were competitors on the scene. It remains to be seen if there will be. In the meantime, ConsumerLab.com’s approach has some positive elements. Founded by a physician formerly with the FDA the group is analyzing various dietary supplements and posting the results on the Internet. It licenses a seal for products that meet its standards. The company hopes to profit by selling its seal to supplement makers and its information to third parties on the Internet, such as health care websites and online retailers.

Unlike the not-for-profit UL, however, ConsumerLab.com is a for-profit (and privately held) operation that currently lacks either competition or government certification, factors that favor efficiency and reliability. There seem to be few safeguards in place to ensure that this new testing company will remain sufficiently independent from the supplement manufacturers with which it seeks to do business. Finally, according to the company’s website, "products are tested on any of the following criteria [emphasis added]": identity and potency, purity, bioavailability, and consistency. Arguably, products should be tested on all of these criteria, but as we have seen, even a perfect score would neglect the fundamental question of whether the active ingredients themselves are safe and effective.

Under any model of nongovernmental regulation that certifies, on a voluntary basis, the safety and efficacy of herbal supplements, manufacturers that choose to participate would gain a measure of protection from liability should a mishap occur, and also support for claims that they make a premium product. Most important, consumers would be assured that the products bearing the Good Housekeeping-like seal of approval had met certain criteria for safety, potency, and quality. They would retain the widest selection of herbal therapies, as well as the freedom to choose certified or noncertified brands.

Perhaps paradoxically, this evolution to greater regulation — albeit voluntary and extragovernmental — would move us closer to a truly free market: consumers exercising independent and informed choice among a broad spectrum of competing products.