Analysis of Over-the-Counter Dietary Supplements

Analysis of Over-the-Counter Dietary Supplements Gary A. Green, MD *; ~Don H. Catlin, MD ; ±Borislav Starcevic, MS *Department of Family Medicine, Division of Sports Medicine, University of California at Los Angeles; ~Department of Molecular and Medical Pharmacology, Department of Medicine, University of California at Los Angeles; and ±UCLA Olympic Analytical Laboratory, Los Angeles, California, U.S.A. CLINICAL JOURNAL OF SPORT MEDICINE 2001;11:254-259 Objective: To determine if steroids containing over-the-counter (OTC) dietary supplements conform to the labeling requirements of the 1994 Dietary Supplement Health and Education Act (DSHEA). Design: 12 brands of OTC supplements containing 8 different steroids were randomly selected for purchase in stores that cater to athletes. There are two androstenediones (4- and 5-androstene-3,17-dione), two androstenediols (4- and 5-androstene-3 , 17 -diol), and 4 more are 19-nor cogeners (19-nor-4- and 5androstene-3,17-dione and 19-nor-4- and 5-androstene-3 , 17 -diol). Main Outcome Measures: 12 brands of OTC anabolic-androgenic supplements were analyzed by high-pressure liquid chromatography. Results: We found that 11 of 12 brands tested did not meet the labeling requirements set out in the 1994 Dietary Supplement Health and Education Act. One brand contained 10 mg of testosterone, a controlled steroid, another contained 77% more than the label stated, and 11 of 12 contained less than the amount stated on the label. Conclusions: These mislabeling problems show that the labels of the dietary steroid supplements studied herein cannot be trusted for content and purity information. In addition, many sport organizations prohibit OTC steroids; thus, athletes who use them are at risk for positive urine test results. In this article we provide the details of the analyses, a summary of the steroids by name and structure, and information on the nature of the positive test results. Athletes and their physicians need this information because of the potential medical consequences and positive urine test results. Key Words: Androstenedione ; 19-Norandrostenedione ; Drug testing ; Sport INTRODUCTION The growth of the nutritional supplement industry in the United States in recent years has been phenomenal. Annual sales are estimated at $12-15 billion, 1 of which $800 million are spent on sport supplements. 2 Critical factors behind this growth are the passage of the 1994 Dietary Supplement Health and Education Act (DSHEA) 3 mandating that U.S. consumers have free access to supplements, the growth of the Internet, and record performances by sports celebrities on supplements. Androstenedione and dehydroepiandrosterone (DHEA), two of the most popular and widely available over-the-counter (OTC) steroids, are metabolic precursors of testosterone, a controlled anabolic-androgenic steroid. Both have been shown to increase plasma levels of testosterone, 4,5 such an increase is a likely prerequisite for enhanced performance. The androstenediones, androstenediols, and DHEA often elevate the testosterone/epitestosterone (T/E) ratio in urine above the threshold of 6:1, which sports organizations consider indicative of suspected testosterone administration. 6 Another group of OTC substances, the 19-norsteroids, result in positive urine test results for 19-norandrosterone, a metabolite of nandrolone (19-nortestosterone). Both testosterone and 19-nortestosterone are schedule III drugs, and virtually all of the scheduled and OTC steroids are prohibited by major sport authorities. Sports organizations have experienced an upsurge in the number of positive urine test results in recent years. In many cases, the athletes claim to have tested positive for compounds that were not listed on the product label. Since 1997, the majority of appeals to the National Collegiate Athletic Association (NCAA) for positive test results have cited the ingestion of sport supplements. Thus there is serious concern over the content and purity of supplements marketed to the athletic community. Likewise, there is concern over the purity, content, and safety of other supplements. 7,8 Consumer groups have tested dietary supplements and found wide variations in the amount of active ingredient in products alleged to be St. John's Wort and echinacea. 1 The analysis of 20 ephedra-containing supplements revealed major discrepancies between the labeled content and the actual content, 9 and some brands of ginseng contained no ginseng. 10 A surprising number of 11,12 Asian herbal medicines are adulterated with undeclared pharmaceuticals. Supplements containing Digitalis lanata , heavy metals, and unidentified contaminants have been linked to a variety of serious disorders, 11,13-15 and 16 recently contamination of androstenedione with trace amounts of 19norandrostenedione has been shown to result in positive urine test results. In the course of analyzing OTC steroids for use in clinical trials, we found significant departures from the ingredients listed on the product label. 16 In the present study we analyzed several brands of OTC steroids to check for compliance with the DSHEA. Specifically, our objective was to determine if these products contained: 1) all of the compounds listed on the label, 2) the correct amount of each ingredient as listed on the label, and 3) any substances not listed on the label. MATERIALS AND METHODS Obtaining Supplements Twelve brands of dietary supplements purporting to contain OTC steroids were randomly selected for purchase in sealed containers from local stores in the Los Angeles, California, area. The 12 brands represented approximately one-half of the brands available to consumers at these retail stores, and one bottle was purchased from each brand. Two to four capsules were selected from each brand for analysis. Chemical Analysis of Supplements Each capsule and its contents were weighed, then the powdered contents were weighed separately and quantitatively transferred to a 100 mL volumetric flask. The empty capsule was weighed and compared with the weight of the full capsule minus the weight of the powder. After the addition of methanol to the flask containing the powder and vortexing, portions of the of the flask contents were further diluted. The amount of steroids in the solutions was determined by high pressure liquid chromatography (HPLC) using a Hewlett Packard 1090 L liquid chromatograph (Hewlett Packard; Palo Alto, CA, U.S.A.) equipped with a diode array detector and a 5 mm Phenomenex (Torrance, CA, U.S.A.) Hypersil BDS 250 × 4.6 mm column. The detector wavelengths were set at 243 and 200 nm. Reference spectra and the retention time (RT) of pure standards of 4-androstene3,17-dione, 4-androstene-3 ,17 -diol, 19-nor-4-androstene-3,17-dione, 19-nor-4androstene-3 ,17 -diol, and testosterone were recorded. The analyses targeted these five steroids. If peaks were found that represented >3% of the main peak, their identity was determined by HPLC and gas chromatography-mass spectrometry. The identity of chromatographic peaks was established by matching the RT and spectrum of each peak to those of the reference standards. Each compound was quantitated against standard curves. In order to set the proper calibration curve range, laboratory technicians were advised of the labeled contents. Categories of Supplements According to the "labeling requirements" in the DSHEA, 3 a dietary supplement is considered misbranded if "the supplement fails: 1) to list the name of each ingredient, 2) to list the quantity of each such ingredient, or 3) to have the identity and strength that the supplement is represented to have." Based on these requirements we devised a system for categorizing the results of the analysis. Because the third criterion for misbranding did not provide a range of acceptable strengths, we adopted the U.S. Pharmacopeia criterion of ±10% for ethical pharmaceuticals. 17 RESULTS Table 1 lists the 12 brands and their respective ingredients as stated exactly on the product label, and provides the average amount found by chemical analysis. To help the reader recognize the steroids, whose chemical names are all similar, the chemical structures and common names of each steroid in Table 1 and related steroids are shown in Figure 1 . The 19-norsteroids, which are shown on the right side of Figure 1 , differ by the absence of a carbon-19 methyl group. 19Norandrosterone, the common metabolite of all the 19-norsteroids, is shown at the bottom. TABLE 1. Ingredients listed on label of 12 brands of steroids compared with testing results FIG. 1. Chemical names, common names, and chemical structures of testosterone and 19-nortestosterone and the family of related steroids that are sold over-thecounter (indicated by OTC). The diols and diones are shown in the middle panel and the common metabolite (19-norandrosterone) of 19-nortestosterone and other 19-norsteroids is shown at the bottom. Additional isomers of the diol steroids exist but have not been found in the OTC market. Three of the brands analyzed (brands 2, 3, and 11) were also reported in another study. 16 Major inconsistencies were found. For example, brand 2 was labeled as containing 250 mg of 4-androstene-3,17-dione (androstenedione) but it actually contained 166 mg with a range of 163-170 mg. Table 1 reveals wider variations between brands than within a particular brand. In 50% of the brands, the differences between capsules taken from the same bottle averaged 10% or less, and the greatest difference was 20%. As shown in Table 2 , category I was defined to include brands in which the average amount of all the steroids present is 90110% of the amount listed on the product label; category II includes brands containing an average of <90% of the amount stated on the label. Category III includes brands containing an ingredient that was not listed on the label. If one of the listed ingredients was absent, the brand was placed in category IV. Category V was used for brands that contained an average of more than 110% of the amount stated on the label. TABLE 2. Classification criteria of over-the-counter steroids Application of these criteria to the results in Table 1 revealed major differences between those brands that were labeled to contain a single ingredient (brands 1-8) and those containing multiple ingredients (brands 9-12). For the single-ingredient brands, only one fell into category I. Of the seven single-ingredient brands in category II, the amount of steroid ranged from 45 to 85%. Although these eight brands (brands 1-8) were labeled to contain only one steroid, brands 1 and 2 contained steroids that were not listed on the label (category III), and in one case that steroid was testosterone. Four brands (brands 9-12) listed more than one ingredient on the label; however, of these, two (brands 9 and 10) did not contain at least one of the steroids listed, and none contained 90-110% of the amount stated on the label. Two of the four multisteroid brands fell into category IV because they did not contain at least one ingredient that was listed on the label. Brand 12 was unusual because the label indicated that it contained 50 mg of a diol steroid, yet we found that it contained an average of 88.5 mg of the steroid. In addition, the label on brand 12 indicated that Tribulus terrestris was present. We did not detect any additional steroids in brand 12, other than the one listed on the label (5-androstene-3 ,17 -diol). However, brand 12 was placed into category V because it exceeded the amount stated on the label, and its capsules actually contained between 166-188% of the announced androstenediol. Administration of OTC steroids listed in Table 1 and others may result in positive urine test results. Table 3 summarizes the positive urine test findings that may result from ingestion of the various steroids. The 19-norsteroids lead to positive results for 19-norandrosterone, the major metabolite of nandrolone (see Figure 1 ). The other steroids often increase the T/E ratio. Androstenedione may result in a positive report for both T/E and 19-norandrosterone. Table 3 also provides literature references for the details of the positive test results. Three supplements that contain stimulants are included in Table 3 to provide a complete listing of other common OTC supplements that may produce positive test results. TABLE 3. Supplements sold over the counter that may result in a positive urine test result DISCUSSION It is abundantly clear from this study that the OTC steroid supplements described herein are mislabeled. Table 2 reveals that seven of the eight single-ingredient and all four of the multiingredient brands tested are "misbranded." The mislabeling covered a variety of errors including listing ingredients that were not found, gross misrepresentation of amount of steroid actually present, and failure to declare steroids not listed on the label. Thus steroid supplements must be added to the growing list of contaminated and mislabeled supplements that are sold to consumers. 7,9,10,12,16,18 An appalling observation on these supplements is that only one brand contained 90-110% of the labeled amount of ingredients. Thus most consumers are receiving substantially less than what is purported on the labels of the supplements tested. One brand contained nearly double the stated amount; therefore, it might trigger adverse reactions in athletes following the manufacturer's guidelines. Another contained testosterone, a prescription anabolic steroid and controlled substance. In some cases, the error was relatively inconsequential: for example, brand 9 lacked three minor ingredients. However, brand 10 that was purported to have 50 mg of 19-norandrostenediol, actually had none. Of course consumers are always advised to read labels, but if the labels are not accurate the consumer cannot make an informed decision. Brand 12 was labeled to contain Tribulus terrestris , a plant containing saponins, diosgenin, and other substances. 19 Diosgenin has been used as a starting material in the industrial synthesis of androgens; however, there is no evidence that these starting materials are metabolized to androgens in man. According to the manufacturers' claims, Tribulus terrestris may increase testosterone levels; however, the only peer-reviewed publication that investigated this issue reported that Tribulus terrestris plus androstenedione had no effect on testosterone levels. 20 Brand 12 contained 5-androstene-3 ,17 -diol as stated on the label and no additional steroids. Due to DSHEA, the Food and Drug Administration (FDA) is not required to monitor the safety, purity, or efficacy of dietary supplements. The labels on supplements may not make health claims. The DSHEA requires supplements to contain all ingredients listed on the label, but there is minimal enforcement. The FDA may act retrospectively, and only if serious toxic events are reported to them. After gamma hydroxy butyrate (GHB) was associated with several adverse events and one death, 21 the FDA did issue an advisory warning. A potential shortcoming of this study is that we studied only 12 brands, thus our findings cannot be generalized to all steroid supplements. Nevertheless there are numerous reports of mislabeled and contaminated OTC steroids. 16 In addition, all the supplements studied herein may cause a positive urine drug test result ( Table 3 ). Administration of androstenedione or androstenediol often elevates the T/E ratio, and the 19-norsteroids are metabolized to 19-norandrosterone. Thus this study adds to the growing body of literature on impurities contained in dietary supplements, and it emphasizes that the OTC steroids or their impurities may lead to positive urine test results. Nowadays many athletes are subject to drug testing. Mislabeling places them at risk for positive test results and being declared ineligible. In general, organizations that conduct athletic drug testing have not accepted the excuse of supplement use to overturn positive results. Therefore it is incumbent upon athletes to consult the appropriate sources of information before taking any OTC steroids. The data reported herein and the summary on positive urine test results indicates that it is not safe for athletes to take any OTC steroids. In addition to the anabolicandrogenic precursors listed in this study, the stimulants ephedrine and caffeine are found in many over-the-counter supplements and are also banned by several sports authorities ( Table 3 ). Many NCAA institutions routinely provide various supplements to their studentathletes. Institutions should be aware that a positive drug test result could result in significant liability for the institution if the supplement is shown to be the origin of the problem. Indeed the NCAA has heard appeals by student-athletes who had tested positive for anabolic steroids and claimed that the only nutritional supplements they had ingested were provided by their school. The NCAA has taken steps to prevent this in that Division II and III schools are now prohibited from distributing dietary supplements. For Division I schools, recent NCAA legislation limits the institutions' ability to provide supplements. While the negative health effects of traditional anabolic-androgenic steroids are well known, reports of adverse reactions associated with OTC steroids are beginning to appear. 22 Federal regulatory agencies are attempting to intervene. For example, the Federal Trade Commission, as part of a settlement with two androstenedione supplement manufacturers (Met-Rx and AST Nutritional Concepts) regarding unsubstantiated safety claims, has ordered that a warning appear on the label: "This product contains steroid hormones that may cause breast enlargement, testicle shrinkage, and infertility in males, and increased facial and body hair, voice deepening and clitoral enlargement in females. Higher doses may increase this risk. If you are at risk for prostate or breast cancer you should not use this product." 23 CONCLUSION The current study validates the concerns of physicians and sporting organizations that the labeling of some "sports" nutritional supplements does not accurately reflect what is contained in the product. This information may be helpful in deterring athletes from using substances that have unsubstantiated efficacy and unknown adverse effects. Acknowledgment: We are grateful to C. K. Hatton, PhD, for scientific and editorial advice, R. Aguilera, PhD, for scientific advice, and the staff at the UCLA Olympic Laboratory for outstanding assistance. REFERENCES 1. Herbal Rx: The promise and pitfalls. Consumer Reports 1999:44-48. 2. Zorpette G. The mystery of muscle. Sci Am 1999; 10: 48- 55. 3. Dietary Supplement Health and Education Act of 1994, Public Law 103-417, 1994. 103rd Congress, 2nd session, S784. 4. Leder BZ, Longcope C, Catlin DH, et al. Oral androstenedione administration and serum testosterone concentrations in young men. JAMA 2000; 283: 779- 782. 5. Mahesh VB, Greenblatt RB. The in vivo conversion of dehydroepiandrosterone and androstenedione to testosterone in the human. Acta Endocrinologica 1962; 41: 400- 406. 6. Bowers LD. Oral dehydroepiandrosterone supplementation can increase the testosterone/epitestosterone (T/E) ratio. Clin Chem 1999; 45: 295- 297. 7. Angell M, Kassirer JP. Alternative medicine-the risks of untested and unregulated remedies. N Engl J Med 1998; 339: 839- 841. 8. Haller CA, Benowitz NL. Adverse cardiovascular and central nervous system events associated with dietary supplements containing ephedra alkaloids. N Engl J Med 2000; 343: 1833- 1838. 9. Gurley BJ, Gardner SF, Hubbard MA. Content versus label claims in ephedra-containing dietary supplements. Am J Health Syst Pharm 2000; 57: 963- 969. 10. Cui J, Garle M, Eneroth P, et al. What do commercial ginsengs contain? Lancet 1994; 344: 134. 11. Huang WF, Wen K-C, Hsiao M-L. Adulteration by synthetic therapeutic substances of traditional Chinese medicines in Taiwan. J Clin Pharmacol 1997; 37: 344- 350. 12. Ko RJ. Adulterants in Asian patent medicines. N Engl J Med 1998; 339: 847. 13. Markowitz SB, Nunez CM, Klitzman S, et al. Lead poisoning due to Hai Ge Fen. The porphyrin content of individual erythrocytes. JAMA 1994; 271: 931- 934. 14. Slifman NR, Obermeyer WR, Aloi BK, et al. Contamination of botanical dietary supplements by Digitalis lanata. N Engl J Med 1998; 339: 806- 810. 15. Williamson BL, Tomlinson AJ, Mishra PK, et al. Structural characterization of contaminants found in commercial preparations of melatonin. Chem Res Toxicol 1998; 11: 234- 240. 16. Catlin DH, Leder BZ, Ahrens B, et al. Trace contamination of over-the-counter androstenedione and positive urine tests for a nandrolone metabolite. JAMA 2000; 284: 2618- 2621. 17. The United States Pharmacopaeial Convention. United States Pharmacopeia. DI 2000, Micromedex, Vol III. Englewood, CO: The United States Pharmacopaeial Convention, Inc., 1998. 18. Anonymous. Melatonin. Med Lett Drugs Ther 1995;37:111-112. 19. Bruneton J. Toxic Plants Dangerous to Humans and Animals. Paris: Lavoisier Tec & Doc, 1999:509-511. 20. Brown GA, Vukovich MD, Reifenrath TA, et al. Effects of anabolic precursors on serum testosterone concentrations and adaptations to resistance training in young men. Inter J Sports Nutr Exer Metab 2000; 10: 340- 359. 21. Center for Disease Control. Multistate outbreak of poisoning associated with illicit use of gamma hydroxy butyrate. MMWR 1990;39:861-863. 22. Kachhi PN, Henderson SO. Priapism after androstenedione intake for athletic performance enhancement. Ann Emerg Med 2000; 35: 391- 393. 23. Federal Trade Commission Files 992 3179-80, Civil Actions #99-WI-2197 and SACV-99-1407. 24. Ros JJ, Pelders MG, De Smet P. A case of positive doping associated with a botanical food supplement. Pharm World Sci 1999; 21: 44- 46. 25. Uralets VP, Gillette PA. Over-the-counter anabolic steroids 4-androsten-3,17-dione; 4-androsten-3 ,17 -diol; and 19-nor-4-androsten-3,17-dione: excretion studies in men. J Anal Toxicol 1999; 23: 357- 366. 26. Uralets VP, Gillette PA. Over-the-counter 5 anabolic steroids 5-androsten-3,17-dione; 5-androsten3 ,17 -diol; dehydroepiandrosterone; and 19-nor-5-androsten-3,17-dione: excretion studies in men. J Anal Toxicol 2000; 24: 188- 193. Received March 2001; accepted September 7, 2001. Address correspondence to Gary A. Green, MD, UCLA Department of Family Medicine, 924 Westwood Blvd., Suite 650, Box 957087, Los Angeles, CA 90095-7087, U.S.A. E-mail: ggreen@mednet.ucla.edu Dr. G. A. Green is Chairman of the NCAA Sub-committee on Drug Testing/Education. Clin J Sport Med 2001 October;11(4):254-259