Estimated daily ingestion levels for the respective metal-containing Ayurvedic medicines were calculated using the mean metal concentration in the product, unit dose weight, and recommended dosage(s) stated on the label. If the manufacturer recommended a range of dosages, a range of potential daily ingestion amounts is shown. India- and US-made products are shown in red and black, respectively. aCalifornia Proposition 65 has established a maximum allowable dose level for lead of 0.5 μg/d.14bAmerican National Standards Institute/National Sanitation Foundation International Dietary Supplement Standard 173 (ANSI 173) states that dietary supplements should not contain undeclared metals that would cause intakes greater than 20 μg/d of lead, 20 μg/d of mercury, and 10 μg/d of arsenic.15cFood and Agricultural Organization/World Health Organization Joint Expert Committee on Food Additives (FAO/WHO JECFA) provisional tolerable weekly intakes correspond to acceptable intakes of 250 μg/d of lead, 50 μg/d of mercury, and 150 μg/d of arsenic for a 70-kg person.17dRasa shastra medicine. eUnited States Environmental Protection Agency reference doses (EPA RfDs) for chronic oral intake of mercuric chloride and arsenic are both 0.3 μg/kg/d, corresponding to 21 μg/d for a 70-kg adult.16
Customize your JAMA Network experience by selecting one or more topics from the list below.
Saper RB, Phillips RS, Sehgal A, et al. Lead, Mercury, and Arsenic in US- and Indian-Manufactured Ayurvedic Medicines Sold via the Internet. JAMA. 2008;300(8):915–923. doi:10.1001/jama.300.8.915
Author Affiliations: Department of Family Medicine, Boston University School of Medicine and Boston Medical Center (Drs Saper and Sehgal and Ms Khouri), Division of General Medicine and Primary Care, Department of Medicine, Beth Israel Deaconess Medical Center (Drs Phillips and Davis), Division for Research and Education in Complementary and Integrative Medical Therapies, Osher Research Center, Harvard Medical School (Drs Phillips and Davis), and Environmental and Occupational Medicine and Epidemiology, Department of Environmental Health, Harvard School of Public Health (Dr Kales), Boston, Massachusetts; New England Regional Laboratory, Environmental Protection Agency, North Chelmsford, Massachusetts (Dr Paquin); National Referral Centre for Lead Poisoning in India and Department of Biochemistry and Biophysics, St John's Medical College, Bangalore (Dr Thuppil); and Employee and Industrial Medicine, Cambridge Health Alliance, Harvard Medical School, Cambridge, Massachusetts (Dr Kales).
Context Lead, mercury, and arsenic have been detected in a substantial proportion of Indian-manufactured traditional Ayurvedic medicines. Metals may be present due to the practice of rasa shastra (combining herbs with metals, minerals, and gems). Whether toxic metals are present in both US- and Indian-manufactured Ayurvedic medicines is unknown.
Objectives To determine the prevalence of Ayurvedic medicines available via the Internet containing detectable lead, mercury, or arsenic and to compare the prevalence of toxic metals in US- vs Indian-manufactured medicines and between rasa shastra and non–rasa shastra medicines.
Design A search using 5 Internet search engines and the search terms Ayurveda and Ayurvedic medicine identified 25 Web sites offering traditional Ayurvedic herbs, formulas, or ingredients commonly used in Ayurveda, indicated for oral use, and available for sale. From 673 identified products, 230 Ayurvedic medicines were randomly selected for purchase in August-October 2005. Country of manufacturer/Web site supplier, rasa shastra status, and claims of Good Manufacturing Practices were recorded. Metal concentrations were measured using x-ray fluorescence spectroscopy.
Main Outcome Measures Prevalence of medicines with detectable toxic metals in the entire sample and stratified by country of manufacture and rasa shastra status.
Results One hundred ninety-three of the 230 requested medicines were received and analyzed. The prevalence of metal-containing products was 20.7% (95% confidence interval [CI], 15.2%-27.1%). The prevalence of metals in US-manufactured products was 21.7% (95% CI, 14.6%-30.4%) compared with 19.5% (95% CI, 11.3%-30.1%) in Indian products (P = .86). Rasa shastra compared with non–rasa shastra medicines had a greater prevalence of metals (40.6% vs 17.1%; P = .007) and higher median concentrations of lead (11.5 μg/g vs 7.0 μg/g; P = .03) and mercury (20 800 μg/g vs 34.5 μg/g; P = .04). Among the metal-containing products, 95% were sold by US Web sites and 75% claimed Good Manufacturing Practices. All metal-containing products exceeded 1 or more standards for acceptable daily intake of toxic metals.
Conclusion One-fifth of both US-manufactured and Indian-manufactured Ayurvedic medicines purchased via the Internet contain detectable lead, mercury, or arsenic.
Ayurveda is a traditional medical system used by a majority of India's 1.1 billion population.1 Ayurveda is also used worldwide by the South Asian diaspora and others.1 However, since 1978 more than 80 cases of lead poisoning associated with Ayurvedic medicine use have been reported worldwide.2,3 Ayurvedic medicines are divided into 2 major types: herbal-only and rasa shastra. Rasa shastra is an ancient practice of deliberately combining herbs with metals (eg, mercury, lead, iron, zinc), minerals (eg, mica), and gems (eg, pearl).4,5 Rasa shastra experts claim that these medicines, if properly prepared and administered, are safe and therapeutic.4,5 Of 70 Ayurvedic medicines manufactured in South Asia and sold in Boston, Massachusetts, stores in 2003, we found that 20% contained lead, mercury, and/or arsenic.6 Estimated daily lead, mercury, and arsenic intakes for these products were all higher than regulatory limits. We identified several rasa shastra medicines that could cause lead and mercury ingestions exceeding US Environmental Protection Agency (EPA) limits by 3 to 4 orders of magnitude. Similar results have been found in other North American cities.7-10
The prevalence of metals in Ayurvedic medicines sold via the Internet and in those manufactured in the United States is unknown. Whether rasa shastra medicines account for most Ayurvedic medicines containing metals and whether they are manufactured by US companies or widely available to US consumers is also unknown.
Thus, this investigation was designed with 3 major objectives: (1) to determine the prevalence of Ayurvedic medicines available via the Internet containing detectable lead, mercury, or arsenic; (2) to compare the prevalence of toxic metals between US- and Indian-manufactured products; and (3) to compare the prevalence of toxic metals in rasa shastra vs non–rasa shastra medicines. We also compared the daily amounts of lead, mercury, and arsenic that would be ingested by persons taking these products according to packaging recommendations with acceptable metal consumption limits suggested by government, industry, and the World Health Organization.
To select Internet-available Ayurvedic medicines for analysis, a strategy used by Morris and Avorn was adapted.11 An Internet search was conducted in November-December 2004 using 5 search engines (Google, Yahoo, AOL, MSN, and Ask Jeeves) and the key words Ayurveda and Ayurvedic medicine. Web sites listed on the first page of results from each search engine were reviewed and all products that met the following inclusion criteria identified: (1) traditional Ayurvedic herbs, formulas, or containing ingredients commonly used in Ayurveda; (2) indicated for oral use; and (3) available for sale. The 673 identified products were entered in a database and a computer-generated random number sequence was used to select 230 products for purchase. Presuming equal numbers of US- and Indian-manufactured products, this sample size provided 90% power to demonstrate a 10% difference in metal prevalence (α = .05). Products were ordered online during August-October 2005. Web site suppliers were not informed of the reason for our purchase.
We recorded the name of each medicine, its manufacturer, and the Web site supplier selling the product. Country of manufacture was defined as the country where the product was formulated (eg, made into a capsule or tablet) and packaged for sale. The country of the Web site supplier was determined from the contact information provided on the Web site. We also recorded formulation, indications, dosage(s), and cost. An Indian-trained Ayurvedic physician (A.S.) classified medicines as rasa shastra if they contained metals, minerals, or gems traditionally used according to 2 classic texts.4,5 Manufacturers claiming Good Manufacturing Practices or metal testing were noted. We determined whether US manufacturers were members of the American Herbal Products Association (AHPA), a trade organization committed to “high-quality herbal products” and “promotion of self-regulation”12 from membership lists available on the association’s Web site. For Indian manufacturers, membership in the Ayurveda Drug Manufacturers' Association (ADMA), committed to “standards and quality amongst its members,”13 was similarly noted. We asked US manufacturers anonymously by telephone where the herbs used for their products were grown. We presumed that herbs used by Indian manufacturers were grown in India. Classification of medicines by these characteristics was done without knowledge of the medicines' metal content.
Products were transferred to plastic vials (Tri-State Distribution, Sparta, Tennessee) with anonymous unique identifiers and transported to the New England Regional EPA laboratory without interruption in chain of custody. From March to October 2006, samples were analyzed in duplicate for lead, mercury, and arsenic concentrations using x-ray fluorescence spectroscopy as previously described.6 The minimum lead, mercury, and arsenic concentrations that could be detected in the medicines with x-ray fluorescence were 5 μg/g, 20 μg/g, and 10 μg/g, respectively. Analysts were blinded to the medicines' rasa shastra status, country of origin, and manufacturer characteristics. Standard reference materials were analyzed alongside the products as positive and negative controls.
Product characteristics were summarized using descriptive statistics. The ratio of products with detectable lead, mercury, and/or arsenic to the number of products analyzed defined the prevalence of metal-containing Ayurvedic medicines in the sample. Ninety-five percent confidence intervals were calculated using the exact binomial distribution. Products were stratified by country of manufacture and the metal prevalences of US- and Indian-manufactured products were compared using the Fisher exact test. Reported metal concentrations are the means of the 2 determinations. In duplicate analysis of 4 products, one lead measurement was below the detection limit of 5 μg/g and the other was at or slightly above the limit; for each of these products the reported lead concentration is the mean of the latter measurement and 0. Median metal concentrations in US- and Indian-manufactured metal-containing products were compared using the Wilcoxon rank sum test. Similar analyses to compare metal prevalence and concentrations between rasa shastra and non–rasa shastra medicines were performed. The Fisher exact test was used to compare characteristics of metal- and non–metal-containing products such as the country of manufacture and Web site supplier, rasa shastra status, manufacturer membership in AHPA and ADMA, Good Manufacturing Practices claims, formulation, herb source, cost, and indication. All analyses were 2-tailed, and P<.05 was considered statistically significant.
Product metal concentration, unit dose weight, and labeled recommended dosage(s) were used to calculate the amounts of lead, mercury, and arsenic that would be ingested on a daily basis if the product was taken as suggested by the manufacturer. These estimates were compared with established acceptable daily limits for metal ingestion. The California Safe Drinking Water and Toxic Enforcement Act (California Proposition 65) has established maximum allowable dose levels for chemicals causing reproductive toxicity (for lead, 0.5 μg/d).14 The American National Standards Institute (ANSI)/National Sanitation Foundation (NSF) International Dietary Supplement Standard 173 is a standard developed by NSF International, a US-based independent, nonprofit, nongovernmental organization, and adopted by ANSI. ANSI 173 provides guidelines to the dietary supplement industry on methods to test and evaluate products, including accuracy of labeling and the presence of specific undeclared contaminants such as toxic metals. ANSI 173 states that dietary supplements should not contain undeclared metals that would cause intakes greater than 20 μg/d of lead, 20 μg/d of mercury, and 10 μg/d of arsenic.15 The EPA has established daily reference doses of 21 μg/d of inorganic mercury and 21 μg/d of inorganic arsenic for a 70-kg adult.16 The Food and Agricultural Organization/World Health Organization Joint Expert Committee on Food Additives establishes provisional tolerable weekly intakes for contaminants in foods of 25 μg/kg of lead, 1.6 μg/kg of mercury, and 15 μg/kg of arsenic, corresponding to acceptable daily intakes of 250 μg/d of lead, 50 μg/d of mercury, and 150 μg/d of arsenic for a 70-kg adult.17
SAS software, version 9.1 (SAS Institute Inc, Cary, North Carolina) was used for all analyses. The Boston University Medical Campus Institutional Review Board exempted the study from human subjects review.
The Internet search identified 25 Web sites featuring 673 Ayurvedic medicines. Of 230 products randomly selected for purchase, we received and analyzed 193 (84%) made by 37 different manufacturers. Reasons for failure to fill our orders included the following: 21 products were no longer available or out of stock; 1 supplier refused to fill our order of 14 products after recognizing that we were authors of a previous study of Ayurvedic medicines6; 1 product received was a duplicate; and 1 product was excluded from study because it was for topical use only. Analyzed and unavailable products did not differ significantly for most characteristics assessed (Table 1).
Overall, the prevalence of Ayurvedic medicines containing detectable lead, mercury, and/or arsenic was 20.7% (Table 2). Lead was the most commonly found metal, followed by mercury and arsenic. The prevalence of metal-containing products did not differ significantly between US- and Indian-manufactured products. The median lead concentration in Indian-manufactured vs US-manufactured lead-containing products was similar. Mercury was present in greater concentrations in Indian-manufactured products. Rasa shastra compared with non–rasa shastra medicines were more than twice as likely to contain metals. Rasa shastra metal-containing medicines had higher lead and mercury median concentrations than non–rasa shastra metal-containing medicines.
The Ayurvedic medicines with detectable metals and their manufacturers, Web site suppliers, rasa shastra status, and metal concentrations are summarized in Table 3. Ayurvedic medicines manufactured in the United States contained primarily lead at concentrations below 25 μg/g, whereas Indian-manufactured medicines contained both lead and mercury at concentrations that reached 104 μg/g. Rasa shastra medicines were more likely to be manufactured in India. Agnitundi Bati, Ekangvir Ras, and Arogyavardhini Bati were Indian-manufactured rasa shastra medicines distributed by US Web sites with extremely elevated lead and mercury concentrations. In contrast, US-manufactured rasa shastra medicines did not contain detectable mercury and had lower lead concentrations than those manufactured in India.
Characteristics of products with and without detectable metals are compared in Table 4. Country of manufacture and Web site distributor did not significantly differ. Manufacturers of 75% of the metal-containing products claimed Good Manufacturing Practices or metal testing, and these claims were not associated with a lower prevalence of toxic metals. Membership in ADMA was not associated with a lower likelihood of metal presence compared with nonmembership. Products made by AHPA members compared with nonmembers were less likely to contain metals. Products containing metals were more likely to be tablets and less likely to be liquids or pastes.
The Figure shows the daily amounts of lead, mercury, and arsenic that would be ingested if metal-containing medicines were taken according to manufacturers' recommendations. Various regulatory standards for maximum daily metal intakes are presented for comparison. All metal-containing products would cause ingestions exceeding at least 1 regulatory standard. Indian-manufactured rasa shastra medicines would cause the greatest lead and mercury ingestions, often substantially exceeding all standards.
In a random sample of commercially prepared Ayurvedic medicines purchased via the Internet, we found that nearly 21% contained detectable levels of lead, mercury, and/or arsenic, and the prevalence of these potentially toxic metals did not differ by country of manufacture; ie, United States vs India. Rasa shastra medicines were more than twice as likely as non–rasa shastra products to contain detectable metals. All metal-containing products exceeded 1 or more standards for acceptable daily metal intake. Several Indian-manufactured rasa shastra medicines could result in lead and/or mercury ingestions 100 to 10 000 times greater than acceptable limits.
Metals identified in our sample of Ayurvedic medicines are likely a result of the practice of rasa shastra or contamination. Many rasa shastra medicines are made with bhasmas, which are elaborately prepared with various forms of metals including cinnabar (mercuric sulfide), galena (lead sulfide), realgar (arsenic sulfide), and white arsenic (arsenic trioxide).4,5 Ekangvir Ras is an example of a rasa shastra medicine made with naga (lead) bhasma and parada (mercury). Ayurveda experts in India believe that if bhasmas are properly prepared according to ancient protocols, the metals undergo shodhana (“purification”), rendering them nontoxic and therapeutic. Case reports in the literature, however, have documented significant toxicity with the use of some of these products.2,3 The prevalence of metals in non–rasa shastra medicines was still substantial (17%) and could be a consequence of environmental contamination of the herbs18 or incidental contamination during manufacturing.
Our finding that close to 21% of Ayurvedic medicines manufactured and distributed by US and Indian companies via the Internet contain lead, mercury, or arsenic is consistent with our previous report that 20% of Indian-manufactured Ayurvedic medicines purchased in South Asian ethnic markets in Boston contain these metals.6 However, the Ayurvedic medicines analyzed in the Boston study had higher median metal concentrations (lead, 40 μg/g; mercury, 20 225 μg/g; and arsenic, 430 μg/g) and were more often recommended for pediatric use (50% vs 5%) than the medicines analyzed in our Internet sample. Studies of Ayurvedic medicines purchased from stores in New York,7 Houston,8 Chicago,9 and Canada10 have also reported similar findings, and lead has been found in non-Ayurvedic herbal and vitamin supplements manufactured in the United States19,20 as well as in traditional medicines from other cultures.21,22
The public health impact of metals in rasa shastra and contaminated herbal medicines in India is unknown and controversial.23 Ayurveda advocates in India maintain that rasa shastra medicines have been used effectively and safely for millennia.4,5,23 They ascribe case reports of metal toxicity to improper commercial manufacturing practices or lack of supervision from a practitioner skilled in rasa shastra.23 However, many Ayurvedic medicine users believed to be unaffected may actually have unrecognized, misdiagnosed, or subclinical metal intoxications. Patients with Ayurvedic medicine–associated lead poisoning commonly undergo endoscopy for abdominal symptoms or bone marrow biopsy for anemia before they receive a correct diagnosis.24 Given widespread use of Ayurvedic medicines in India and throughout the world, observational studies assessing whether rasa shastra and non–rasa shastra medicine use are independent risk factors for increased lead burden are urgently needed.
Limitations of our study include potential misclassification of the product's country of manufacture and rasa shastra status. Information provided by the Web site, label, and/or manufacturer was occasionally contradictory or ambiguous. However, this limitation does not affect the overall prevalence of metals in the sampled medicines, and any misclassification would have been nondifferential. Unobtainable products were not random and may have had a higher or lower likelihood of containing metals, thus potentially affecting the overall sample prevalence. We did not assess batch-to-batch variability in metal concentrations. However, this would likely not significantly alter overall prevalence estimates. Wide variation among published standards for acceptable limits of daily metal ingestion makes it difficult to assess the magnitude of potential toxicity for different products. Finally, we did not ascertain the specific physical form or chemical species of the metals. To the best of our knowledge, the physicochemical form of metals in rasa shastra medicines and their bioavailability have not been fully characterized or reported. We are unaware of rigorous evidence supporting claims that bhasmas made using lead, mercury, and arsenic are nontoxic, and documented case reports of poisonings2,3 contradict these theories.
Regarding generalizability, products in our study can be purchased via the Internet without consultation from an Ayurvedic practitioner. Thus, our results may not reflect products recommended or provided by individual Ayurvedic practitioners to patients in the context of a patient-practitioner consultation.25 Compared with Ayurvedic practitioners in India, US practitioners are reportedly less likely to use rasa shastra medicines.25 Rasa shastra is not included in the scope of practice being developed by the US-based National Ayurvedic Medical Association, a professional organization for US- and Indian-trained Ayurvedic practitioners (Jennifer Rioux, PhD, personal communication, 2008). Internet products may also not be similar to those sold over the counter in mainstream US pharmacies and health food stores.
Despite these limitations, our data likely provide a relatively accurate snapshot of the prevalence of metal-containing Ayurvedic medicines sold via the Internet in 2005. A 2005 Institute of Medicine report concluded that “the regulatory mechanisms for monitoring the safety of dietary supplements . . . [should] be revised. The constraints imposed on FDA [US Food and Drug Administration] with regard to ensuring the absence of unreasonable risk associated with the use of dietary supplements make it difficult for the health of the American public to be adequately protected.”26 A 2006 National Institutes of Health state-of-the-science conference similarly stated that “public assurance of the safety and quality of [dietary supplements] is inadequate.”27 Our data demonstrating elevated levels of lead, mercury, and arsenic in publicly available Ayurvedic medicines and prior studies finding toxic metals in non-Ayurvedic supplements19-22 support these conclusions. New FDA regulations28,29 and current Indian policies23,30 do not specify any maximum acceptable concentrations or daily dose limits for metals in dietary supplements for domestic use. We suggest strictly enforced, government-mandated daily dose limits for toxic metals in all dietary supplements and requirements that all manufacturers demonstrate compliance through independent third-party testing.
Corresponding Author: Robert B. Saper, MD, MPH, Department of Family Medicine, Boston Medical Center, One Boston Medical Center Place, Dowling 5 S, Boston, MA 02118-2317 (firstname.lastname@example.org).
Author Contributions: Dr Saper had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Saper, Kales.
Acquisition of data: Saper, Khouri, Paquin, Thuppil.
Analysis and interpretation of data: Saper, Phillips, Sehgal, Davis, Kales.
Drafting of the manuscript: Saper.
Critical revision of the manuscript for important intellectual content: Saper, Phillips, Sehgal, Khouri, Davis, Paquin, Thuppil, Kales.
Statistical analysis: Saper, Davis.
Obtained funding: Saper.
Administrative, technical, or material support: Saper, Khouri, Kales.
Study supervision: Saper, Phillips.
Financial Disclosures: In 2004, prior to the beginning of this study, Dr Sehgal was employed for 1 year as a research associate by Arya Vaidya Pharmacy, an Indian-based Ayurvedic medicine manufacturer. No products from Arya Vaidya Pharmacy were analyzed in this study. No other disclosures were reported.
Funding/Support: Dr Saper is supported by a Career Development Award (K07 AT002915-03) from the National Center for Complementary and Alternative Medicine (NCCAM), National Institutes of Health. Dr Phillips is supported by a Mid-Career Investigator Award (5K24AT000589-07) from NCCAM.
Role of the Sponsor: The Centers for Disease Control Childhood Lead Poisoning Prevention Program (CDC CLPPP) was provided with a list of Ayurvedic medicines to purchase for the study. The CDC CLPPP ordered the medicines with its funds. The CDC CLPPP also provided funds to the New England Regional EPA Laboratory to defray the cost of analyzing the medicines. Otherwise, the CDC LPPP and NCCAM had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or preparation, review, or approval of the manuscript for submission.
Disclaimer: The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health, NCCAM, EPA, CDC, or National Referral Centre for Lead Poisoning in India.
Additional Contributions: We gratefully acknowledge Jennifer Rioux, PhD (University of North Carolina Department of Anthropology), Larry Culpepper, MD, MPH (Department of Family Medicine, Boston Medical Center), and Karen Freund, MD, MPH (Department of Medicine, Boston Medical Center), for helpful reviews of earlier versions of the manuscript, for which they were not compensated. We also thank Florence Uzogara, MA, and Surya Karri, MBBS, MPH (Boston Medical Center, compensated by Department of Family Medicine funds), for research assistance.