The science and regulations of probiotic food and supplement product labeling

Authors


Address for correspondence: Mary Ellen Sanders, Ph.D., Dairy and Food Culture Technologies, 7119 S. Glencoe Ct., Centennial, CO 80122. mes@mesanders.com; or Dan D. Levy, Ph.D., Microbiologist, Supervisor, New Dietary Ingredient Review Team, Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration, 5100 Paint Branch Parkway, College Park, MD 20740. Dan.Levy@fda.hhs.gov

Abstract

Presented by the New York Academy of Sciences, the U.S. Food and Drug Administration (FDA), and the U.S. Office of Dietary Supplements of the National Institutes of Health, the symposium “Probiotic Foods and Supplements: The Science and Regulations of Labeling,” was held on June 12, 2010 at the New York Academy of Sciences, New York, NY, the goals of which were to facilitate the exchange of ideas regarding labeling and substantiation of claims for probiotics among academic, industry, and regulatory professionals, and to discuss ways to translate and communicate research results in a truthful way to the consumer and to such health professionals as physicians, pharmacists, and dieticians. The target audience for this symposium included academicians interested in conducting research on the health benefits of probiotics; scientists; communications personnel, and regulatory specialists from companies involved in, or interested in, the marketing of probiotics; U.S. government regulatory experts tasked with oversight of probiotic foods and dietary supplement products; and other experts in the field interested in the development of probiotics for the U.S. market.

Introduction

This conference focused on issues related to probiotic product labeling. Dan Levy from the FDA gave an introduction to the conference by discussing the interactions of science with regulatory requirements for food labels. Regulatory requirements such as laws and regulations, while authoritative, tend to be less detailed than the scientific base for the information that appears on the food label. Tools that connect the authoritative rules with the detailed science include guidance documents published by the FDA and other regulatory agencies but also standards set by private organizations or industry trade organizations. Standards set by private organizations, while not reviewed or endorsed by government, are often influential. In addition, there is always the option of formal or informal consultations between industry stakeholders and regulators. The goals for the sessions were laid out: to explore how the science can be used within the regulatory framework and to produce a meeting report that promotes labeling that conforms to regulatory requirements and accurately informs consumers. Issues specifically addressed during the conference included the labeling of probiotic conventional foods and dietary supplements with statements of beneficial health effects as well as the labeling of ingredient content. A “provocative theme question for the day” was also introduced: “Is the term probiotic, when used on a food label, an embedded structure/function claim for which substantiation should be furnished?”

Session 1: Structure/function claims for probiotic conventional foods or dietary supplements

This first session was chaired by Mary Ellen Sanders (Dairy and Food Culture Technologies). It dealt with labeling as it pertains to the communication of structure/function claims for conventional foods or dietary supplements. Structure/function claims relate consumption of the product to the normal functioning of the human body.1 They are distinguished from health claims, which describe a relationship between a whole food, food component, or dietary supplement ingredient, and reducing risk of a disease or health-related condition, and drug claims, which refer to an article's use in the diagnosis, cure, mitigation, treatment, or prevention of disease. Drugs are further defined as articles (other than food) intended to affect the structure or any function of the body.2 A common probiotic structure/function claim, “digestive health,” was chosen as an example for the discussions.

This session was designed to bring together several relevant perspectives in considering the issue of labeling of probiotic conventional foods and dietary supplements with health benefit claims. The perspectives represented included those of a practicing gastroenterologist, a clinical researcher, a specialist in consumer trends and opinions on health benefit claims, regulatory authorities (Food and Drug Administration's Center for Food Safety and Applied Nutrition and the Federal Trade Commission (FTC)), and a representative from a center within the National Institutes of Health, the National Center for Complementary and Alternative Medicine (NCCAM).

This forum provided a unique opportunity for researchers, practitioners, industry, and federal government representatives to engage in a dialogue on some of the challenges facing the probiotic industry. Constructing claim language is viewed by some in the industry as a type of linguistic gymnastics. For example, the claim of “helps with sexual potency” would be a drug claim, whereas “improves sexual performance” would be a structure/function claim (Heimbach, J., personal communication), inasmuch as impotency is a disease, poor performance is not. Most companies doing business in this field are interested in marketing their probiotics as conventional foods or dietary supplements. However, their intent is often for these products to provide dietary approaches to help consumers cope with health conditions, such as mild irritable bowel syndrome, or reducing the risk of getting common infectious diseases or travelers’ diarrhea. Although such uses are typical ones evaluated for probiotic use, claims regarding these uses are generally perceived by the U.S. regulatory agencies as not acceptable for conventional foods or dietary supplements.

The probiotic community at the meeting expressed a need to better understand U.S. regulatory agency expectations of marketers of probiotic products with regard to what types of information would constitute adequate substantiation of claims on probiotic products. In addition, a better understanding of appropriate uses of conventional foods or supplements for targets affected by probiotics was sought.

Approach for substantiating a claim of digestive health for probiotics

Yehuda Ringel (University of North Carolina at Chapel Hill) set the stage with perspectives from points of view evident to him as a practicing gastroenterologist and clinical researcher on claims of digestive health for probiotics.

  • 1Patients’ perspectives: there is clearly great interest by patients in probiotics. Most are convinced that food is a trigger for their functional bowel symptoms, and they look to their physicians for guidance on using probiotics.
  • 2Physicians’ perspectives: physicians are aware of emerging research on probiotics but are put in a bind when governmental and professional organizations, such as the FDA, NIH, U.S. Department of Agriculture, American Gastroenterological Association, and American College of Gastroenterology, refer patients to physicians for guidance on use of probiotics and other types of dietary supplements without publishing evidence-based guidelines to help physicians with their recommendations to patients.2-4 The practicing physician is in a position to respond to patients’ questions on the use of probiotics to relieve and prevent common digestive problems, such as constipation, bloating, abdominal discomfort, and diarrhea. Although research suggests that certain probiotics (at the right dose and in the right formulation) can help restore the proper balance of the intestinal microbiota and possibly improve gastrointestinal symptoms,5,6 available data have not been translated into practice guidelines to aid physicians in making proper, evidence-based recommendations.7 Furthermore, many of the probiotic products currently available on the market likely have not been scientifically tested for specific clinical uses. Physicians need more data on currently available probiotic products and guidance on how to advise patients on the use of probiotics.
  • 3Clinical researchers' perspectives: from the clinical research point of view, there are three key questions that must be considered: the condition being tested, the study population, and the study endpoints. Regarding the condition to be tested or measured, there is a discrepancy between the clinical meaning of such terms as health,8 disease, symptom, syndrome,9 or illness, and the regulatory authorities’ and non-regulatory federal agencies’ view of these terms as they relate to food and drug law.Medical definitions of these terms can be found in Figure 1. Ringel shared his perspective that although it may be justified that the FDA is requiring an Investigational New Drug (IND) application for conditions that are associated with known diseases, an IND may not be needed for physiologic disruptions, symptoms, and syndromes that are functional in nature and are not considered diseases. The study population—and how it relates to the target population—was also discussed. The example of subjects presenting with functional bowel symptoms provides a good example. The general population comprises people who are healthy but have gastrointestinal (GI) symptoms, and both those who have abnormal GI functions but are not seeking medical care (non-patients) and those who are seeking care (patients). Most people with functional GI symptoms do not seek medical care.10 There are differences between non-patients and patients with functional bowel symptoms.11 Non-patients have fewer days with pain; lower severity of pain, distention, and bloating; less dissatisfaction with bowel habits; a better quality of life; and less time that their symptoms interfere with their everyday life. Ringel stated that although non-patients may not be considered “healthy,” they should be considered part of the general population. Regarding study endpoints, in the context of the topic of the session, there is no medical definition of the term digestive health, and therefore as such, this cannot be an endpoint for a clinical study. Numerous gut functions can be measured, for example, motility, transit, sensation, digestion, absorption, fermentation, and immune functions. But an important issue to resolve before choosing one of these physiological endpoints for a study is the extent to which measured changes in these physiological parameters relate to clinical response. Therefore, it is necessary to understand the clinical relevance of the physiological effect being measured.
  • 4Personal perspectives: Ringel emphasized that additional research is needed on the content, safety, benefits, and mechanisms of action of available probiotics. The expertise exists to conduct the needed high-quality studies. However, obstacles delaying this progress are lack of resources as well as different opinions regarding the appropriate focus for research on probiotics, conventional foods, and supplements among parties involved in the process, such as the FDA, the NIH, industry, and academic researchers (Fig. 2). Fueling this disagreement may be discrepancies in endpoint terminology among marketing professionals, researchers, healthcare providers, regulatory authorities, and consumers.
Figure 1.

Medical definitions of terms used to describe states of health and disease. The terms health and disease are insufficient to describe the continuum of states experienced by the population.

Figure 2.

Summary of Ringel's perspectives on the state of the science of probiotics as it applies to clinical use and research.

Probiotics and label claims: communicating the health benefits of foods that can promote health

Elizabeth Rahavi (International Food Information Council (IFIC)), provided perspectives on how consumers view health benefit claims on food labels, with a special emphasis on probiotic and prebiotic foods and their relationship to digestive health. Surveys conducted by the IFIC12-14 provide evidence that consumer awareness of diet and health relationships is at the highest level since the survey began in 2000. This is possibly due to upward trends in media reports that convey the benefits of foods, beverages, and food components; Figure 3 shows the upward trends of media coverage specifically for probiotics and prebiotics.

Figure 3.

Media coverage of probiotic and prebiotic stories, 2000–2010.

The surveys also show that the majority of consumers (91%) believe they have some control over their health. Furthermore, they believe that food and nutrition play the greatest role—more than exercise or family health history—in maintaining their health. This suggests that consumers fundamentally agree with the concept of “functional foods,” in that 89% strongly or somewhat agree that food plays a role in health beyond basic nutrition. Additional findings from the 2009 IFIC Foundation Food and Health Survey14 show that, indeed, the majority of consumers (64%) report making changes to improve the healthfulness of their diet. Improving overall well-being, improving physical health, and losing weight are the top reasons why people are making dietary changes.14 Additionally, more than eight out of ten consumers are interested in learning more about foods and beverages with added health benefits.

Rahavi considered the consumer perspectives on digestive health. For many consumers, digestive health is a state of being, and food can have an immediate effect on an individual's digestive health.13 If consumers find relief by consuming a product such as a probiotic, then they are more likely to continue that behavior in the future, which may encourage continued consumption of those specific foods over time. The Functional Foods/Foods for Health Survey13 indicated that nearly three quarters of Americans are aware of the putative health benefits of probiotics for maintaining healthy digestive and immune systems, a significant increase from 2007. However, consumers weren't the only ones showing increased awareness of probiotics and prebiotics. For the first time, the 2010 Dietary Guidelines Advisory Committee (DGAC) evaluated the role of probiotics and prebiotics in health. While the experts believed that gut microbiota play a role in health, they did not feel there was adequate basis for making specific recommendations about probiotics or prebiotics. As indicated in this report, “Thus, the DGAC believes that the gut microbiota do play a role in health, although the research in this area is still developing. No recommendations for intake of prebiotics or probiotics for Americans can be made, although foods high in prebiotics (wheat, onions, garlic) should be consumed, as well as food containing probiotic (yogurt).” Another observation from the IFIC Foundation Food and Health Survey14 was that although 61% of consumers look at food labels for information to guide their food and health practices, only 20% say that they are looking for information about health benefits. Additional focus group research related to health claims, which was conducted by the IFIC Foundation in 2005, found that structure/function claims were perhaps the most popular of all claims tested.12 Most participants preferred the simpler, clearer, and more easily understood structure/function claim messages over qualified health claims.

Many factors and influences, not just labeling, were shown to affect consumers’ food choices. Price, taste, and health benefits are all part of the “choice” equation. But the challenge for the future appears to be how to motivate consumers to sustain beneficial dietary changes. Simply being aware of benefits is not sufficient for consumers to implement dietary changes. There is a great need for effective communication that is consistent and accurate and delivers positive messages (tells what to do, not what not to do), which empowers consumers with knowledge to make healthy choices.15 With regard to probiotics and prebiotics, a critical element in these communication efforts is developing messages that accurately reflect demonstrated health benefits without overstating their benefits.

What does the FDA consider adequate substantiation of structure/function claims?

Kenneth M. Taylor (Center for Food Safety and Applied Nutrition, United States Food and Drug Administration) discussed the FDA's guidance for substantiating structure/function statements that are made for dietary supplements as provided by 21 U.S.C. 343 (r)(6), as published in the 2009 guidance document, “Guidance for Industry: Substantiation for Dietary Supplement Claims Made Under Section 403(r) (6) of the Federal Food, Drug, and Cosmetic Act.”16 Although the FDA has not established any legally enforceable responsibilities, it has drawn upon its own expertise with regulations and case law, Federal Trade Commission (FTC) experience and policy on dietary supplement claims made in advertising, and recommendations from the Commission on Dietary Supplement Labels17 to develop guidance that describes the amount, type, and quality of supporting evidence that may be considered appropriate or acceptable. The FDA's approach applies a substantiation standard of “competent and reliable scientific evidence” to determine if a claim is adequately supported. This substantiation standard, which is consistent with FTC case law, defines a standard of competent and reliable scientific evidence as “tests, analyses, research, studies, or other evidence based on the expertise of professionals in the relevant area, that has been conducted and evaluated in an objective manner by persons qualified to do so, using procedures generally accepted in the profession to yield accurate and reliable results.” Because of the broad nature of this substantiation standard, there is no preestablished formula as to how many or the types of studies that may be needed to substantiate a claim, although the guidance does give examples of the types of information that would be needed to substantiate 21 claims given as examples.

However, when applying a substantiation standard of “competent and reliable scientific evidence,” relevant areas, such as the meaning of the claim(s), the relationship of evidence to the claim(s), the quality of the evidence, and the totality of the evidence, should be considered when assessing if a structure/function claim is adequately supported within the context of a surrounding body of evidence. Therefore, to determine the quantity and type of appropriate studies that would be necessary for substantiating a claim, the FDA will consider the accepted norms in consultation with experts from various disciplines, or, if available, an existing standard for substantiation by either a government agency or other authoritative body.

A claim's meaning or intent must first be understood by identifying each implied and expressed claim. In situations where a claim may have more than one reasonable meaning, then the firm should have substantiation for each interpretation. Firms should not focus exclusively on individual statements, but also on the overall message when all statements are collectively considered. Consumer testing may be useful to determine the ordinary common understanding of claims in context. For example, the claim “Recommended by Scientists” implies that there is a body of qualified experts who believe that the claim is supported by valid evidence. Consumers might also reasonably interpret this statement to mean that there is general scientific agreement or consensus regarding the claim. So if the evidence does not demonstrate consensus, or is the opinion of a single scientist or a small group, then the claim probably does not have sufficient substantiation.

The relationship of studies or evidence to a claim is another important consideration for adequate substantiation. To illustrate, studies used to support a claim should be similar in formulation, serving size, length, and frequency of exposure, as well as be conducted under conditions as those indicated on the labeling of the actual product. Manufacturers should account for the presence of other substances, whether naturally present in the study as artifacts or in the subject product, that might affect the dietary supplement's performance or the study results themselves, and if the study used an appropriate population. The studies should also specify measured endpoints used to substantiate the claimed effect. Finally, the manufacturers should evaluate if the claim conveys the extent, nature, or permanence of the effect and the level of scientific certainty.

Scientific quality is another consideration when deciding if studies adequately substantiate a claim. Scientific quality is based upon several criteria, including (1) the study population, including factors such as number, age, gender, and health; (2) how the study is designed and conducted, use of placebo controls, and data collection methods; (3) statistical analysis; and (4) outcome measures. If a scientific study adequately addresses most or all of these criteria, then it may be considered to be high quality. While a randomized, double-blind, placebo-controlled trial design is ideal for achieving excellent scientific quality, such trials may not always be possible, practical, or ethical. However, reliable scientific evidence that is adequate to substantiate a claim will consist of information primarily derived from human studies, and generalizing among different populations may not be scientifically valid. Some information has limited usefulness to substantiate dietary supplement claims. Animal and in vitro studies may provide useful background information on the biological effects of a substance, but often have limited or unknown value in predicting the effect in humans. Care should be taken when extrapolating animal research directly to human condition. Anecdotal evidence such as consumer testimonials generally is not sufficient to substantiate dietary supplement claims because each individual's experience might be attributable to factors other than the dietary supplement itself. Evidence, such as review articles and product monographs, may provide background information that is useful to understand the scientific issues about the relationship between the dietary supplement and the claimed effect. Multiple factors can affect a study's design, and these should be considered when determining the reliability of the study. Potential sources of bias include lack of appropriate randomization and blind controls, the number of subjects, demographics, and the reproducibility of results. Confounders include variability in dosage quantity or the presence of other substances that may have independent effects. Factors that may contribute to study reliability include quality assessment criteria; for example, the study population should be assessed for size and for the extent to which it represents the target population for the product claim. Assessment of intervention or exposure outcomes, which addresses parameters like defined and appropriately measured dosing and efforts to detect harmful as well as beneficial effects, tend to support improving study quality. And finally, although not required, publication in peer-reviewed journals also adds some level of assurance that qualified experts have reviewed the research and found it to be of sufficient quality and validity; however, the mere fact that a study happens to be published is not necessarily an indicator that the research is competent and reliable evidence to adequately substantiate a particular claim.

When determining if there is adequate evidence to substantiate a claim, evidence should be considered in its entirety. This includes considering all relevant research, whether favorable or unfavorable. The evidence used to substantiate a claim should be consistent with the surrounding total body of evidence. As conflicting or inconsistent results can undermine the validity of a claim, plausible explanations should be attempted to reconcile discrepancies. Disagreement among studies could be attributed to variables, such as concentration, assay methods, or study populations.

The FDA has established a uniform approach to evaluating supporting evidence for substantiating structure/function claims with a substantiation standard. This standard provides flexibility for considering a variety of supporting evidence that may be used. Having quality evidence to adequately support structure/function claims increases consumer confidence in the product, including novel ones like probiotics when used in dietary supplements.

Substantiation of health claims in advertising

Richard L. Cleland (Federal Trade Commission) discussed how the federal regulation of messages to consumers regarding probiotic products is divided between the FTC and FDA. The FTC has primary jurisdiction over the advertising of probiotic products, while the FDA has primary jurisdiction over the labeling. However, because both agencies use very broad definitions of labeling and advertising, there is considerable overlapping jurisdiction. The FTC Act prohibits “unfair or deceptive acts or practices in or affecting commerce” and the “false advertisement” of any food, drug, device, service, or cosmetic. An advertisement is deceptive if it contains a representation or omission of fact that is likely to mislead a consumer acting reasonably under the circumstances, and that is material to a consumer's purchasing decision. In general, advertisements containing false or unsubstantiated claims, or that fail to disclose material facts, are deceptive. This section will focus broadly on health-related claims.

When reviewing advertisements, the first step in the analysis is determining meaning. The FTC looks at both expressed and implied claims. Expressed claims speak for themselves. For implied claims, the FTC uses a “reasonable consumer” standard to determine the “net impression” of the advertisement. Where the meaning of an advertisement is reasonably apparent on its face, extrinsic evidence of its meaning is not required. Advertisements may convey more than one meaning. To be deceptive, an implied claim need not be the only, or even the predominant, message.

Regulatory requirements under the Food, Drug, and Cosmetic Act depend, in large part, on the classification of the product, for example, drug, food, or cosmetic that, in turn, can depend on intended use. In contrast, the FTC's substantiation analysis is claims driven, and the inquiry focuses on whether the available scientific evidence is sufficient to support the advertising claims. In other words, the amount of substantiation required for a claim about boosting the immune system would not vary depending on whether the product was considered a drug or food. In addition, the FTC does not classify studies into drug and non-drug studies.

All express and reasonably implied material claims conveyed by an advertisement must be truthful and non-misleading. Making objective health-related claims about a product or service in an advertisement without a reasonable basis is a deceptive practice and constitutes false advertising. If an advertisement represents that a claim is supported by a particular level of evidence, for example, clinically proven, the advertiser must possess and rely on at least that level of evidence to substantiate the claim.

In general, health-related claims for products containing probiotics require competent and reliable scientific evidence to substantiate that the representation is true. The evidence is considered in light of the entire body of relevant and reliable scientific evidence, not just the evidence supporting the representation. There is no limitation in the FTC Act to presenting or discussing supportive published literature on an Internet website that is being used to promote the sale of a product as long as the expressed and implied claims being made for the product through reference to that literature are substantiated. Moreover, as a general rule, in such situations, there is no affirmative obligation to disclose non-supportive research. However, linking to scientific papers that address a probiotic's ability to treat or prevent a disease or health-related condition may be considered to be making a claim for the product that must be substantiated by competent and reliable scientific evidence.

The FTC recognizes that claims may be qualified but cautions that in practice such claims may be hard to make without confusing consumers. Claims based on emerging science must be truthful and non-misleading. To constitute reliable substantiation, a study must be conducted in a manner that allows for valid extrapolation from the study population to the target population of the advertisement. The FTC's standard is flexible but rigorous. Some common flaws in studies submitted to the FTC include (1) relying on studies of different ingredients, in different doses, and in different forms of administration; (2) lack of adequate controls; (3) insufficient or lost data; (4) no statistical analysis or inappropriate statistical analysis; and (5) the use of invalid or unvalidated measurements.

The fact that there may be responders and non-responders in a study does not change the basic scientific requirements. However, a marketer can tailor its advertising to subpopulations so long as the advertiser's qualification is effective and there is research that adequately supports the claimed benefit for the targeted population. Where advertisers often get into trouble is taking studies done on subpopulations, for example, persons deficient in certain nutrients, and then trying to extrapolate the findings to a larger population. Of course, post hoc manipulation of responding populations within a study is neither scientifically sound nor convincing evidence for the FTC.

The quantity and quality of evidence required to support the claim “supports digestive health” would depend on how consumers interpret that claim. As is often the case with ambiguous claims, the consumer's understanding is likely to depend on the context in which the claim is made. Extrinsic evidence would also be important. For example, if consumer surveys showed that a significant number of consumers took away an implied claim that the product would reduce transit time, then the advertiser would have to have well-controlled human clinical studies demonstrating such an effect. Given the saturation of “support” claims in today's functional food and dietary supplement market, it is hard to tell what kind of secondary meanings the word “support” has assumed. At a minimum, however, the advertiser must establish some level of benefit related to digestion and then substantiate that the product provides that benefit. Here, also, the quantity and quality of evidence required depends on the proposed benefit.

Probiotic claims are becoming more popular with marketers. Probiotic claims made for disease treatment or prevention are likely to continue to be an agency priority.

Panel discussion

Neville Bamji (Mount Sinai Medical Center)

Richard L. Cleland (Federal Trade Commission)

Linda Duffy (National Center for Complementary and Alternative Medicine, NIH)

Elizabeth B. Rahavi (International Food Information Council)

Yehuda Ringel (University of North Carolina at Chapel Hill)

Mary Ellen Sanders (Dairy and Food Culture Technologies)

Kenneth M. Taylor (U.S. Food and Drug Administration)

The chair opened the floor for questions from the audience, although discussion questions were developed to spur the panel discussion if needed (Table 1). The audience clearly expressed general concern about regulatory approaches to probiotic research and how it can be used to substantiate structure/function claims.

Table 1.  List of questions presented for the discussion on Structure/Function claim substantiation. Note that some of these claims might be appropriate as “health claims” (see text).a
  1. aMany of the questions posed were not addressed during the discussion because of time limitations.

Claim substantiation
 • To what extent can changes in gut microbe populations or metabolic activity (or other surrogate endpoints or biomarkers) be used to substantiate probiotic claims?
 • What is your opinion of establishing an independent system to evaluate if a health-benefit claim is substantiated, that is similar to generally recognized as safe (GRAS), but focused on expert opinions of efficacy (not safety)?
 • What information is needed to substantiate a structure/function claim on “supports digestive health” that meets the standard of “truthful and not misleading”?
Scope of allowable claims for foods/supplements
 • Are these diseases or health conditions?
   ○ improper balance of microbiota
   ○ mild IBS symptoms
   ○ reduction of the risk of acute, but diet-related conditions, such as colds or allergic symptoms
 • What is the harm in allowing true, scientifically supported, health-benefit claims for foods or supplements that go beyond the current FDA judgment of food vs. disease? Is there a down side?
 • Is there an implied health claim when the word immunity is used in claims, even when part of a properly worded structure/function claim?
 • Does the current system of restricting probiotic claims to structure/function claims serve as an effective mechanism to promote safety and efficacy of probiotics?
 • The restriction of using only structure/function claims on foods and supplements provides a disservice to patients and physicians. The general nature of these claims gives a false impression that probiotics are the same. What is the justification for limiting the incorporation of rigorous science to support a clinical indication?
 • Currently, foods are not considered for dietary management of symptoms or for preventing an acute disease or condition in a healthy or at-risk population (such as preventing colds or diarrhea). This seems to be a limited view of what food can do, seems to run counter to what physicians recommend (for example, many recommend yogurt when giving people antibiotic prescriptions, recommend prunes or high fiber foods for constipation, etc.), seems to ignore what the U.S. population is seeking to understand, and seems to shackle U.S. researchers.
Claim elasticity
 • Can a blend of probiotic strains use clinical results from an individual strain that is a component of the blend?
 • Can a clinical study of a probiotic conducted on one delivery format be used to substantiate claims for delivery of the same probiotic strain/dose in a different format?
 • Probiotics must be formulated with greater numbers than are indicated on the label, and greater numbers than tested in clinical studies, to meet a minimum count at the end of shelf life. Is this OK?
 • Can a probiotic blend use clinical results from an individual strain?

One concern was with how much of the spectrum between health and disease is categorized as disease for regulatory purposes. Researchers would like to test the ability of conventional foods or dietary supplements to help people with their health challenges, but the FDA appears to consider most of these endpoints as “treatment” or “prevention” of disease and thus appropriate for drugs, not foods. This theme also emerged as a concern of medical professionals during Ringel's presentation during this session. There was an extensive discussion of this topic at a related New York Academy of Sciences conference.18

This led to a request for guidance from the FDA on how to substantiate health benefits of probiotics. On the one hand, it was clear from the lectures that both the FTC and the FDA expect all meanings of a claim to be substantiated, but research targeted at addressing what consumers expect from probiotics, such as reduction of risk of acute diseases or health conditions, or dietary management of disease symptoms, hasn't seemed to be seen by the FDA as appropriate for conventional foods or dietary supplements. Attempts by industry to substantiate benefits of probiotics and provide meaningful, physiological results for consumers and healthcare providers run the risk of having the probiotic deemed a drug by the FDA. Guidance on how to design a study that is physiologically meaningful but does not trigger drug status is needed. Concern was expressed that scientific investigation into effects and mechanisms is hindered by the FDA's insistence for an IND for studies of products that are never intended by manufacturers to be marketed as drugs or those for which the ultimate positioning may not be clear. If the proposed research use is deemed safe, the value of insisting on INDs for studies of products marketed as conventional foods or dietary supplements was not evident to many in the audience.

It was pointed out by an audience participant that research endpoints are not product label claims. Although the FDA has legal jurisdiction over product labeling claims, the participant questioned their authority over research study endpoints. Insisting that probiotics be studied under INDs was characterized as a restriction of the research that is needed (and sought by medical professionals) to understand the mechanisms of action and the extent of effects that probiotics might have. Although an endpoint in a study may be interpreted by some in the FDA to be a drug-use endpoint, this doesn't mean that the product being tested will be marketed or used as a drug. So while certain studies could potentially contribute greatly to understanding the role of the test product in health, and could therefore provide useful scientific knowledge as part of a dossier for efficacy, researchers should be aware that in certain circumstances, from a regulatory standpoint, a study can be considered to make an implicit disease claim depending on how it is used in marketing a product. Researchers who conduct studies involving disease-related matters run the risk, assuming the product is not already a dietary ingredient, of having the product become a new drug, depending on the extent and nature of the research and if it becomes recognized for its drug effects. Taylor pointed out that research can become part of a product's labeling if it is referenced through marketing of a product. He was unable to discuss whether Center for Food Safety and Applied Nutrition (CFSAN) was planning to provide guidance on this issue, and representatives from the FDA's Center for Biologics Evaluation and Research, the Center that oversees INDs for probiotics, were not present on the panel to respond to concerns about INDs.

A direct question to Taylor was posed: If one were interested in developing scientific substantiation for a “supports immune function” structure/function claim for a conventional food or dietary supplement, would it be reasonable to do a study with an endpoint of reducing the risk of colds, along with immune biomarkers? He responded: “Colds are considered diseases. It would probably be inappropriate to use a study which examines a disease condition to substantiate a structure/function claim. Manufacturers would risk that such a study to substantiate a ‘supports immune function’ claim could be taken by ordinary consumers to imply a therapeutic benefit for colds and subject their product to regulatory enforcement action if it is deemed violative.” According to the FDA guidance, if the study population is healthy at the initiation of the study, then information on the benefits of a supplement on the immune biomarkers could be used to substantiate a structure/function claim for a dietary supplement. By contrast, if the study population already had colds, then the data for the effect of the dietary supplement on the immune biomarkers would not be suitable to substantiate a structure/function immunity claim in healthy consumers because the study did not use healthy subjects. Therefore, if a reasonable consumer is going to interpret a claim to mean that the product is beneficial in the treatment, mitigation, or prevention of a disease, then the statement is most likely a disease claim and subject to regulatory action. If this is the case for “supports immune function,” then this claim would not be allowable on a conventional food or supplement. Whether such a claim could be made following a petition for a health claim regulation for a food was not discussed.

Duffy referenced a joint committee report19,20 in which recommendations were made by an expert committee of food scientists, microbiologists, and nutritionists to enhance the validity of labeling and health claims of probiotic-containing food products. She raised the question posed by the committee report that asked whether regulatory conceptualization of health benefit claims might be expanded. The discussion then focused on the particular concern about an immunity claim that implies a health benefit beyond just maintaining one's immune system. The current research and regulatory perspective is that such claims need to be supported by reliable scientific evidence of a clinically substantiated enhancement of the health benefit. A related issue was raised regarding the role of foods containing probiotics to complement drug therapy by providing support for, or tolerance of, the drug.

The other members of the panel were asked whether they considered the suggested expanded conceptualization a reasonable and achievable approach. In response, Taylor indicated that if there is interest to have a broadened framework considered, one possible approach is usually through a citizen's petition, which is a request for the FDA to take an administrative action, such as modifying an existing regulation or promulgating a new regulation.

The use by companies and the FDA of expert panels to provide opinion on research issues such as substantiation of claims was questioned. One problem with expert panels is that their conclusions are wholly dependent on their composition, and it is not rare that two panels comprising highly qualified experts from different specialties come to different conclusions when considering the same data. It is problematic if a company relies on its own panel of experts, but the FDA relies on a different panel that reaches different conclusions. Taylor commented that the courts have given deference to the FDA and FTC as regulatory agencies when different interpretations of law and science may each have merit. The FDA and FTC have the prerogative to assemble experts to reach a conclusion that will enable it to make the most informed decision. Cleland emphasized that an expert's opinion is never better than the data on which it is based.

Rahavi was asked to comment on the extent to which consumers believed health benefit claims. She indicated that the IFIC surveys did not survey for “belief.” However, she noted in a study of consumer understanding of difference in the grades of evidence for qualified health claims, consumers had trouble understanding the different levels of evidence, and in fact preferred the wording of structure/function claims, which they found more “motivating.”21

Regarding the need for strain-specific studies for probiotics, Cleland indicated that, based on the current state of understanding of probiotics, results from studies on one strain could not be extrapolated to other strains. However, a scientific rationale for such extrapolation could be considered on a case-by-case basis.

A question regarding appropriate research approaches for different types of claims was raised. From the FTC's perspective, a specific claim such as “prevents colds and flu” would require a higher standard of evidence for substantiation than a claim for “supporting immune function.” Biomarker data may be enough to substantiate a “supports immune function” claim. The FTC would rely on experts in the field to help them judge, but two double-blinded, placebo-controlled studies might not be necessary for such a general claim. In addition, the FTC does not enforce the Food, Drug, and Cosmetic Act and will not directly examine whether a product labeled as a food is really a drug because the claim changes the intended use, for example, to prevent a disease, only whether the claim is substantiated. Taylor added from the FDA perspective that a manufacturer intending to market a product to prevent the cold or flu is not within the permitted intended use for a dietary supplement, but agreed that for an appropriate nutrition support statement regarding immunity, something less than a double-blinded, placebo-controlled study can be sufficient provided that the supporting evidence meets the substantiation standard of competent and reliable scientific evidence.

A question regarding an apparent gap that exists between science and claims was posed from the audience. This gap reflects difficulty in translating the jargon and complexity of scientific findings into product and advertising claim language. A balance must be struck between scientific accuracy and claims that are correctly understood by the consumer. The current FTC standard is that claims must be truthful and not misleading as understood by a “reasonable consumer.” Taylor pointed out that the FDA and the scientific experts and available case law on which it may rely would have to agree that the wording of the marketing claim is indeed supported by the available evidence. The question arose whether the FDA and FTC would consider well-conducted, advertiser-sponsored consumer research that showed how consumers understood the claims. Both Cleland and Taylor indicated that such consumer research, if properly designed, would be relevant and useful to understanding whether a claim was substantiated. However, Cleland cautioned that companies must substantiate all reasonable interpretations of a claim. Under precedents established by extensive case law, a claim understood in a certain way, even by a substantial minority (10–15%) of consumers, must still be substantiated. This concern could be very relevant with the example of digestive health. Since there is no scientific definition of this claim and since consumers are likely to have a myriad of different digestive concerns, this claim could mean many different things to consumers. Furthermore, the intention of the advertiser does not matter from the FTC perspective. What matters is how consumers are interpreting the claim.

Ringel was asked if he believed any probiotic products could legitimately make a claim for promoting digestive health. After answering yes, he indicated that in his opinion, such general claims do a disservice to the well-conducted studies published on some probiotics. These studies have endpoints much more specific than the generic term digestive health. Such a claim masks the true findings of the study. As discussed already, digestive health is an imprecise term that is not really meaningful to healthcare professionals and probably not to consumers as well. It would be better if study endpoints could be more directly communicated, but he acknowledged the regulatory barriers for this.

Taylor stated that the FDA has not addressed the term digestive health in either the structure/function regulation or the claims substantiation guidance. Therefore, such claims will be evaluated on a case-by-case basis within the context of how the claim is made.

Cleland and Rahavi were asked how effective the FTC has been at exercising its authority. Cleland responded that it is not equipped to police the entire marketplace and they have to do the best they can with limited resources.

A complicated question was asked about how the science can meet the regulatory standard. Ringel suggested that a multipart discussion with the science community, regulatory community, and consumers may advance understanding of this issue. He suggested that there is a need for more research on probiotic benefits. But he added that once these results are in hand, a discussion will be needed with regulatory agencies about how the results fit the regulatory paradigm. Furthermore, discussions with consumers about their needs and, more importantly, how they understand information being given to them are necessary. He suggested that meetings such as this one are needed to help resolve these concerns, pointing out that this was one of the first such attempts to do so.

Reflecting continued discussion from the previous day (the Probiotics: From Bench to Market conference), a question was asked regarding the difficulty of getting data to support benefits of conventional foods and supplements within the context of requirements for INDs. Considerable confusion and discontent was expressed over when a human study must be conducted under an IND, especially since some study endpoints would not be used as label claims. Study endpoints could include some that look like the intent is to treat disease, but the spectrum of endpoints could be wider than the ultimate intended label claim. Taylor responded that if the study population is sick (e.g., people with colds being studied to support a claim of boosting the immune system), it is difficult to apply the results to healthy individuals, and so it is not evident how the study will support anything other than a drug claim. On the other hand, conditions like occasional diarrhea and constipation are not diseases and are legitimate endpoints for structure/function claims. The questioner suggested that there was a minefield for clinical researchers since the intended use of the marketed product may be different from the tested product.

In conclusion, this discussion presented a unique opportunity for researchers, practitioners, and industry and regulatory representatives to discuss some of the challenges facing the probiotic industry. The evolving nature of the science behind the impact of probiotics on human physiology was acknowledged, as was the promise that this field may hold, as further understanding of the role of microbes in health and disease is elucidated. No disagreement was expressed over the need for all health benefit claims, including structure/function claims, to be substantiated for probiotics. But concern was expressed by stakeholders over the options that are acceptable within the current regulatory framework for human study endpoints upon which to build a case for substantiation of claims for food products. This regulatory framework includes food law and regulatory agency interpretation of that law. A product that reduces the risk of disease is allowable for foods in the United States, but no petitions have been submitted for a regulation allowing this type of claim for a probiotic ingredient. In Europe, as in the United States, claims about reduction of risk—but not prevention or treatment—of acute disease, such as colds or flu, are allowable on conventional foods, although there has been much debate about probiotic claims on foods, and decisions by the European Food Standards Agency have proved controversial. Industry stakeholders present at the meeting felt that a broadening of the scope of health benefit claims for food and supplements to encompass health endpoints relevant to probiotics would impel both industry and academic research in the field, as well as allow product labels to reflect specific, rather than general, benefits. Although not allowed under the current regulatory structure, stakeholders expressed interest in broadening the scope of claims acceptable on foods. Stakeholders thought that targets such as improved tolerance to medical therapies, improved therapeutic effects of drug therapy, dietary management of refractory symptoms, and impact on acute health conditions are conceivable roles for foods, if such uses meet the safety standards for foods. Broadening the scope of allowable claims in this manner would require acceptance of a much broader role for conventional foods and dietary supplements in maintaining health than envisioned by the legal definition of a drug as any article intended to diagnose, treat, cure, mitigate, or prevent disease. The current definition channels products with these intended uses into the drug development pipeline, requiring the products to meet the safety and efficacy standards for drugs, standards that require considerably higher expense and longer development time.

Session 2: Nomenclature for probiotic microorganisms and strains

Geraldine June (the Food and Drug Administration) introduced the regulatory requirements for food labels. The FDA regulates food labeling under a variety of statutes, the most important being the Federal Food, Drug, and Cosmetic Act. Statutes are enforced by regulations codified in Title 21 of the Code of Federal Regulations.

There is no FDA regulatory definition for the term probiotic. Foods—whether they are dietary supplements or conventional foods—must be labeled in specific ways. The label must contain mandatory elements, such as the statement of identity (the name of the food), ingredient statement, amount of food in the package, and other material facts about the food, as required by regulation (e.g., allergen warnings). The overarching requirement is that all labeling must be “truthful and not misleading.”

The statement of identity is the standard of identity name, if there is one. The standard of identity can be defined by regulations (e.g., yogurt at 21 CFR 131.200). If there is no standard of identity, the common or usual name or a descriptive name is used, unless the nature of the food is obvious, in which case a “fanciful” name is permitted. General principles for common or usual names are that they should accurately identify the basic nature or properties of the food, be uniform among similar products, and not be confusingly similar to other foods not reasonably encompassed by the same name. Coined terms can be used when appropriate, frequently after consultations with the FDA. Examples of coined terms include acceptance of “canola oil” for “low euricic acid rapeseed oil” and “orange roughy” for “slimehead fish.” In some cases, there are regulations establishing common or usual names, such as for peanut spread (21 CFR 102.23). The FDA has also given guidance on appropriate common or usual names, as in the seafood list.22 Common or usual name regulations are less prescriptive than standard of identity regulations. The FDA can also issue guidance or letters stating opinions about common or usual names and also consults informally with companies by telephone and email to discuss names. When there is more than one ingredient, the names of the ingredients must be declared by common or usual name in the label in descending order of predominance by weight. In general, ingredient names must be specific, not collective names, although there are certain exemptions, including permission to use “cultured” to refer to bacterial cultures used to culture certain products such as “cultured skim milk.”

Rules for dietary supplements are slightly different than those for conventional foods. Dietary ingredients listed in the supplement facts panel on the label of a dietary supplement need not be repeated in the ingredient list, for example. Botanicals must be identified by the part of the plant from which they are derived and by their the common or usual name, which must be consistent with the names standardized in Herbs of Commerce,23 if the botanicals are included in that reference—otherwise, the Latin binomial name, as determined by the International Code of Botanical Nomenclature,24 must be used. There are no explicit regulations for naming bacteria used as ingredients in conventional foods or dietary supplements. Common or usual names should be consistent with the general principles for common or usual names and should be consistent throughout the industry. Scientific names for bacteria should be consistent with the regulations that describe how to apply scientific names for botanicals. The principles for establishing scientific names for bacterial ingredients are the subject of the remainder of this session. The common or usual name of the bacterium should be used unless an alternative name has been established by regulations, guidance, or by consultation with the FDA. Finally, the name on the ingredient list should be the same as the name used if the bacterium were to be sold elsewhere as such.

During the discussion, an audience member asked whether an organism cultured using food allergens in the culture media should be declared on the food label. In response, June reviewed the general criteria for an exemption: the manufacturer must demonstrate, in a petition or a notification to the FDA, that the food either does not contain any protein or does not contain the allergen. Details about petitions and notifications under the Food Allergen Labeling and Consumer Protection Act (FALCPA) are available on the FDA website.25 A second question was asked about the appropriate units of measure for viable organisms, particularly as to whether the colony-forming unit (CFU) is acceptable. In response, June pointed out that there are no regulations or guidance documents specific to this question, so there is no formal policy. Levy noted that in responding to New Dietary Ingredient (NDI) notifications, the FDA has been requesting information by weight and an appropriate range of CFU per serving under the assumption that there will be some change over the course of shelf-life.

Charles Franz (Max Rubner Institute, Federal Institute for Nutrition and Food, Germany, and member of the Lactobacillus Committee of the International Committee on Systematic Bacteriology) discussed the taxonomy of lactobacilli and bacilli. While he was unable to find any regulations, recommendations from a variety of international bodies (e.g., Health Canada, EFSA, and an FAO/WHO working group) urge the use of the Latin binomial name and strain to designate probiotic organisms used as ingredients. Nomenclature for bacteria is one of three subdivisions of the science of systematics. The rules for nomenclature are established under the International Code of Nomenclature of Bacteria, which is kept up to date by the Judicial Commission of the International Committee on Systematic Bacteriology. A genus is loosely defined as a collection of species with many shared characteristics and a limit of genetic divergence of between 10–12 percent. Species is defined for eukaryotic organisms based on the ability to reproduce and bear fertile offspring. The concept of species is more difficult to define for prokaryotes since they are haploid and reproduce asexually. Species identification for prokaryotes relies on a combination of overall similarity (monophyletic and genomically coherent clusters of organisms), as well as diagnosis by a discriminative phenotype, for example, genetic similarity >97% based on the 16S RNA sequence. The code also recognizes subspecies as minor but consistent phenotypic or genetically distinct clusters within a species. The International Code of Nomenclature includes General Considerations, Principles, Rules, and Recommendations that extend to subspecies nomenclature. However, this system has no rules or recommendations below the level of subspecies (e.g., for strains).

Most organisms used as probiotics are in one of two phyla, Firmicutes or Actinobacteria. Firmicutes contains both Lactobacillaceae and Enterococcaceae families in the Lactobacilliales order or the Bacilliales family in the Bacillales order. Bifidobacteria, also used as probiotics, are in Actinobacteria. The genus Lactobacillus contains over 130 validly published species names and is unusually large and genetically diverse. For example, the GC content of the DNA varies from 32–54%, twice the acceptable range, suggesting that the taxonomy is not yet well established and likely subject to further revision. 16S ribosomal DNA sequencing suggests five main branches, each of which might be a separate genus. While there are some correlations using this technique (for example, groups A and E seem to cluster into organisms of gut origin, and groups C and D are likely to be of food origin), there is no consistent biological correlation for the groupings, for example, when compared to a tree built using DNA sequences using protein coding sequences instead of ribosomal sequences.26 In fact, one group identified by this latter method seems closer to E. faecalis than to other lactobacilli. An example of the state of flux for lactobacillus nomenclature over the last 40 years was described: the L. casei type strain turned out to be particularly unrepresentative of the species, and many related strains previously called paracasei, rhamnosus, or zeae are now considered part of the species. Bifidobacteria are more coherent, but nonetheless there have been nomenclature shifts among the 31 currently valid species names, such as the consolidation of B. lactis as a subspecies of B. animalis. Similarly B. infantis, B. longum, and B. suis are now considered types of a single species now designated as Bifidobacteria longum. Bacillus illustrates an additional challenge for probiotics since it not only is overly diverse as a single genus, but also combines pathogens such as B. cereus with safe strains with a long history of use in fermented foods. Misclassification on product labels has been reported in the literature, such as products labeled as containing B. subtilis that actually contain organisms from other species, such as the enterotoxiogenic B. cereus and B. clausii.27,28

In summary, current genus and species designations should be used on labels. The use of out-of-date names is not always wrong, but it can be misleading in specific cases (e.g., the use of Streptococcus faecium to describe what is correctly called Enterococcus faecium). Given the dynamic nature of nomenclature rules, one has to expect shifts from time to time and should consult authoritative sources, such as the lists of valid names published in the International Journal of Systematic and Evolutionary Microbiology and reproduced online at J.P. Euzeby, List of Prokaryotic Names (http://www.bacterio.cict.fr/) and DSM Bacterial Nomenclature Up-To-Date (http://www.dsmz.de/microorganisms/bacterial_nomenclature). The use of strain and trade names is not subject to taxonomic rules of nomenclature and should not be misleading. The deposit of strains in an international culture collection, as well as proper storage and use of master cultures, should be employed to preserve the integrity of the strain (Fig. 4).

Figure 4.

Phylogeny of the Fermicutes, which contains most of the species investigated for probiotic properties.

In response to questions from the audience, Franz pointed out that subtypes (as opposed to strains) are subject to nomenclature rules and, to be correct, should be included in the name. He also acknowledged that type strains are sometimes atypical of the majority of the strains in the species, but maintained that they remain useful as a common basis for comparison for any laboratory to use and thus remain useful in spite of their limitations. The prospect of breaking up lactobacillus into multiple new genera was particularly worrisome from the perspective of a company worried about confusing consumers by changes in genus name on product labels, although Franz suggested that there seems to be no prospect for this happening in the near future.

Brian Beck (American Type Culture Collection (ATCC)) described ATCC and its role as an international biological resource center (BRC) for identifying and culturing bacterial strains for commercial use. ATCC provides research and development tools and reagents as well as related biological material management services and is an independent, nonprofit entity. Its many roles and services include authentication of materials submitted for deposit, preservation of cultures, and providing methods of distribution compliant with the requirements of regulatory bodies and the depositor.

ATCC offers many custom services, including production services, such as creation of master cell banks and identification services specific to the organism. Uniquely, ATCC has also received ISO accreditation for many of its systems. Commercially important strains are accepted into the general collection with no fee to the depositor, but recipients are charged a fee. The general collection contains many types of strains, including the type strains for many species. Specialized collections include contract collections, such as for the NIH Human Microbiome Project, and the patent collection. For a fee, deposits are accepted into the patent collection as mandated under international patent rules, and the service includes a notification to the depositor when the material is requested. A safe deposit collection is a nonpublic collection used by manufacturers for commercial purposes like a master cell bank or other archive.

Strain processing generally starts with testing for viability and is propagated into archive and distribution lots. It is then authenticated using “polyphasic testing”: morphological, physiological, and molecular typing as appropriate to achieve an accurate identification for the microorganism. Techniques include traditional culture, plating, and biochemical testing as well as molecular methods and “next-gen” DNA sequencing, high-throughput phenotypic arrays, and mass spectroscopy–based metabonomics, which can be applied as appropriate. 16S sequence information is generally the first step but is insufficient by itself and must be complemented by additional methods to achieve satisfactory resolution down to the strain level. ATCC works with the depositor to determine the appropriate taxonomic identity at the time of deposit, and can work with them after the deposit to keep up with evolving rules of nomenclature.

During the question period, Beck was asked whether ATCC routinely accepts strains being sold as food ingredients into the general collection. He responded that while there are limits to what ATCC will accept into the general collection, strains that are being used for commercial purposes are likely to be accepted, although depositors may be asked to provide evidence that there will be sufficient demand for samples from the collection once deposited to justify placement in the general collection.

Eric Brown (CFSAN/FDA) discussed the concept of “strain” and the challenge of identifying the boundaries of the designation of strain in the face of steady evolutionary shifts in bacteria. His perspective is based on his experience in investigating and tracking pathogens in the FDA foods program rather than probiotic organisms intentionally added as food ingredients. There is no standard definition of a strain, although the concept of a variant with a significant difference (e.g., H1N1 influenza or MRSA) is familiar to both microbiologists and the general public. Many systems designate strain differences based on the property illuminated by a particular type of test (e.g., serovar, morphofar, pathovar). These tests can be related to important phenoypes (e.g., virulence, acid resistance) or biomarkers less directly related to the organism's physiology (e.g., serovar or designation of geographical information about where first collected). “Type strains” provide a useful anchor against which other members of species can be compared, but there is no expectation that the type strain is necessarily “representative” of the entire species, although when it is not representative that does introduce confusion. There are no standard rules for strain designation, and there are many methods that have been used to identify differences among various strains.

Pulse field gel electrophoreses (PFGE) is frequently the first method used to investigate the relationships among a collection of related organisms. Digestion of the bacterial genome using sequence-specific restriction enzymes generates a two-dimensional pattern of bands on an electrophoresis gel. While this technique can differentiate closely related species from one another (e.g., Salmonella typhimurium from S. enteriditis) and can distinguish and differentiate among dozens of strains within those species, it is not always fine enough to differentiate down to epidemiologically relevant levels in outbreak investigations. DNA sequence-based methods, such as optical sequencing, Diversalab, and whole-genome sequencing (WGS), provide more detailed information. WGS is now fast enough (three days) and inexpensive enough ($1000) to be used routinely in outbreak investigations. The data can be searched to identify single nucleotide polymorphisms (SNPs) or, in an additional few days of effort, annotated as to phenotype, such as by connecting the genetic changes to large databases of genetic information, such as the National Center of Biotechnology Information at the National Library of Medicine and using tools such as the “COG” analysis (http://www.ncbi.nlm.nih.gov/COG/) to assess the significance of the genetic changes. These analyses can reveal important information about the physiological significance of seemingly small and subtle shifts in a DNA sequence. For example, in an outbreak of Salmonella enteriditis, WGS was instrumental in linking outbreak isolates to the food source. In addition, strains recovered from patients were tightly clustered within the outbreak cluster, and the shifts that created the patient cluster were located in genes associated with increased survival in the human GI tract, strongly suggesting that these strains were able to rapidly accumulate during a single passage through humans, traits that better adapted them to persistence in the human GI tract. Thus, organisms that live in a variety of environments can quickly evolve to exploit new environments, and the use of sophisticated high-density sequencing tools can identify these differences. Another example given was acquisition of the Lac+ phenotype, not usually associated with Salmonella. Significantly, this outbreak investigation discovered evidence that this survival trait was able to transfer horizontally to other Salmonellae and spread quickly through the gene pool. Thus, a variety of techniques are used to characterize isolates. The most detailed of these methods can generate information with maximum diversity, unique to each particular isolate. The challenge is thus interpreting this information to allow placement of isolates into clusters of related organisms we call strains in ways that are useful and meaningful toward understanding the relevant biological information, as well as using strain nomenclature that accurately identifies organisms as new strains once a significant variation has occurred (Fig. 6).

Figure 6.

Investigation of a Salmonella outbreak showed rapid divergence of strains once ingested.

Questions addressed to Eric Brown revolved around how much sequence variation is needed to be a significant change. For example, one audience member asked whether the generation of a small number of SNPs is enough to require a change of strain designation, since that can be inevitably generated during routine passage. He responded that while you can insist that a single SNP generates a new strain, it is more practical, particularly in the context of this meeting, to regard these as re-isolates of the same strain unless there is a detectable phenotypic difference. Another audience member suggested WGS as part of routine quality assurance during production or, as an alternative, to develop chip-based technologies focused on genetic targets for significant phenotypic changes. While Dr. Brown acknowledged that chip-based technologies (as well as other advanced genomic, proteomic, and metabolomic diagnostics being developed) are clearly where things are heading, he suggested that currently it can be satisfactory to use fewer data-dense alternatives, such as barcoding or PFGE.

Duane Charbonneau (Procter and Gamble) discussed the naming and marketing of strains from an industrial perspective. Companies must navigate among various needs, including regulatory requirements, quality assurance during production, and customer assurance, and that there is an understandable connection between the material in the product and the test materials used in clinical testing. A commercial enterprise is confronted with an explosion of information and of tools proposed for use in bacterial identification. They use a variety of phenotypic and genotypic methods that include multilocus enzyme electrophoreses, protein fingerprinting by gel electrophoreses, and classical tests like biochemical tests for substrate utilization and genetic methods, such as ribotyping and other 16S RNA–based methods, multi-locus sequence typing, and various restriction fragment and PCR-based methods. Newer methods, such as time-of-flight mass spectroscopy–based proteomics and next-generation WGS and transcriptomic methods, while not yet routine, are being developed for more common use. Systems biology, particularly focused on host/organism interactions, will be key tools for understanding these organisms. However, we are still in the early stages of these new technologies, and there is not yet enough public data to use them effectively. For example, of the over 4,000 whole genomes in the public database, only a handful are lactobacilli. There is not enough depth yet for the types of analyses that are possible with other organisms better represented in the database. The Diversalab Bacterial Barcodes system is an example of the use of a technique—in this case one that generates a two-dimensional pattern of bands on a gel created by PCR amplification of products generated using sequences of repetitive sequences scattered through bacterial genomes in patterns that help identify specific strains. The Latin binomial genus and species names are used to identify ingredients. These are the most useful names in that they provide standardized information about the organism, and there are rules of nomenclature. Strain names are more problematic, because while they are useful in uniquely identifying the ingredient, they do not always clearly convey information about the organism. Perhaps this is an area where more guidance is needed. Trade names are a useful anchor to provide a clear message to consumers, especially given the instability of taxonomic nomenclature, which can change over time as standards for nomenclature shift. On the other hand, while trade names are a popular way to be reflective of the organism, they must also be consistent with regulatory requirements, which can vary based on the regulatory category of the product. There is also some tension between scientists, who develop the organisms, and marketing people, who are often responsible for decisions about trade names. The scientists do have some responsibility to make sure that the trade name is not misleading. Shifts in nomenclature are inevitable, and as more information is learned about an organism, a new name can be beneficial. A process is needed that helps all stakeholders connect a particular strain to previously published literature and regulatory submissions that used a different name. The label of a consumer product should accurately identify the organisms using the scientific genus and species names as well as have a strain designation, although trade names are necessary additions to help consumers understand the nature of the ingredient, particularly when there have been shifts in taxonomic rules. These rule changes are an inevitable part of shifting conventions and evolving standards. Guidance is needed to help industry react appropriately when such changes occur (Fig. 7).

Figure 7.

Discussion session

The discussion started with a proposal to attempt to answer four questions:

  • 1What are the proper boundaries on designating species and strain names in the list of ingredients for a food product?
  • 2How can the name of an organism change (e.g., to keep up with new conventions of nomenclature) without misleading consumers about the nature or “newness” of the ingredient?
  • 3To what extent is it appropriate and useful to have fanciful, trade, or “common and usual” names, and how do they relate to the scientific name for the purpose of labeling?
  • 4What are the advantages and disadvantages of using a name that contains an implied claim or phenotype? Should the substantiating evidence be publicly available?

There was much discussion on how to identify organisms on the food label. For example, does the history of using the generic term cultures for cultured products make this term appropriate for products using strains with probiotic properties? There are many published recommendations, such as from an expert panel convened by FAO/WHO in 2002,29 that each organism in a product be labeled by species and strain. But is this necessary? Ms. June responded that the common and usual name of a bacterium is clearly its genus and species name and, if there are properties that are significant to the product and unique to the strain, then the strain name and/or an appropriate trade name probably becomes necessary. Most fanciful names for ingredients, like “peas” and “corn,” came into use before the plants were identified scientifically. In contrast, organisms being used in probiotic foods are known by their scientific names prior to being used in food, so a more compelling rationale would have to be used to allow substitution of a fanciful name for the scientific one. However, because of the novelty of this issue, there probably should be some sort of interaction between the FDA and stakeholders to try to achieve a consensus on the best way to implement these considerations.

An audience member pointed out the difference between a cultured product for which the organisms transform the food by fermenting the initial ingredients, and a probiotic food in which the organism is intended to benefit the consumer directly and may or may not ferment the food. Charbonneau, responding to an inquiry about whether a trade name alone was adequate to identify a mixture of strains, suggested that to the extent that the identity and quantity of each strain is needed to meet a claim, each of these is a material fact that should be disclosed on the label. Charles Franz gave an example of how use of a trade name helped avoid confusion when a species name changed. Biogarde is a trade name familiar to consumers in Germany for a popular yogurt product. One of the three strains used in it changed nomenclature, but consumers relied on the product trade name, although in this case the trade name applied to the food product, not the ingredient. Levy posed the following question: In the United States, dietary supplement law requires a new premarket safety notification to the FDA when an ingredient is chemically altered. How much change from one strain to the next is sufficient to make the new strain chemically altered? This generated vigorous discussion. Charbonneau suggested that one use “today's tools to get today's answer,” understanding that standards will shift with improved technology. Brown suggested that the key question is whether the change is phenotypically significant. In particular, does it alter the phenotype of the strain in ways that alter the safety profile or ability to meet any claim?

An audience member pointed out that there is no a priori way of knowing whether the change is significant. Since it is unreasonable to repeat all information collected during the initial development of the product, one must ask what is a reasonable minimum set of tests to characterize the new variant. Another audience member pointed out that rather than burdening a small number of individual companies by asking them to address this problem, industry must work with regulators and academics to develop standards applicable industry-wide. Suggestions from the audience included public hearings conducted by the FDA and that voluntary standards could be developed by non-governmental organizations such as AOAC. Franz pointed out that in Europe, at least for safety analyses (as opposed to efficacy) for those species listed as having a qualified presumption of safety (QPS) status, this is not an issue since that standard presumes any strain in the species is safe and suitable for use in food.

The next discussion topic, identification of the quantity of the organisms in the product, started with a discussion of an appropriate “overage” to account for declines in viability over the course of shelf life. Audience suggestions included the caveat that the maximum amount in the product must be consistent with the safety data; that under dietary supplement good manufacturing practice, variations in viability must be thought out by the manufacturer as part of the development of the manufacturing process; and that for conventional foods, the GRAS program at the FDA has a history of requesting this information in notifications. Also, because expiration dates and survival can be affected by temperature and other storage conditions, labeling should be clear about proper storage needed to maintain product quality. It is also unclear whether the quantity should be reported as a single number (e.g., the maximum amount) or as a range. For conventional foods, there is no requirement for amounts or expiration dates. For supplements, however, the quantity in the supplements facts panel must be met but only while the product is in interstate commerce, which does not include the period after the consumer has purchased the product.

The use of weight in addition or instead of CFU to identify the quantity of the ingredient was unpopular with many stakeholders since it can vary with the manufacturing technology used, and while it might give some idea of the ratio of live to dead cells, it also can include excipients, spent medium, and other non-bacterial components, which can legitimately vary considerably from batch to batch. While it might not be the best measure of quantity for all purposes, the weight is required to be listed for each dietary ingredient in dietary supplements and is required to be used to determine the order of ingredients on a conventional food label. However, proprietary blends that do not disclose the quantity of each ingredient are permitted for dietary supplements, so any standard that requires quantitative information for each strain might have to be promulgated by an industry stakeholder organization since it may not be possible for the FDA to impose this as a requirement, although it can enforce a label claim that a product meets an independent standard. In addition, nothing prevents a company from voluntarily putting truthful and non-misleading information about the contents on the label. The standard is whether the label is misleading. This is determined in part to see whether the label is misleading without the information, and whether consumers would be misled about the product in the absence of this information.

The session ended with a discussion of the question introduced at the beginning of the day: Is “probiotic” a structure/function claim for which substantiation is necessary? No consensus was reached, although the question generated a vigorous discussion. Some audience members defended the practice of using probiotic only in a very general way without any requirement for substantiation. By contrast, others pointed out that the FAO/WHO consult definition does not tell you what the health benefit is, only that there is one. It was also pointed out the FAO/WHO consult is only a recommendation and is not universally accepted. The question is whether this statement is “puffery” (promotional statements and claims that express subjective rather than objective views, such that no reasonable person would take them literally). The FDA and FTC perspective is to ask whether consumers are being misled. There did seem to be a consensus that there needs to be considerable latitude about the nature of the health benefit that would permit an ingredient to be labeled as a probiotic. There was also a suggestion from the audience that the rules and expectations for conventional foods might be somewhat different than for dietary supplements. Some concern was expressed that a health benefit could be interpreted to mean a health claim, which has specific regulatory implications; however, by the end of the discussion, it was clear that health benefit did not automatically imply health claim as defined in labeling regulations.

Overall recommendations and conclusions

  • 1Bacteria used as probiotic ingredients should be identified by the genus and species name that conforms to current scientific standards for bacterial nomenclature. An appropriate strain name should also be used, but not one that implies an unsubstantiated claim.
  • 2Strain and trade names are appropriate ways to identify organisms, particularly when bacterial nomenclature rules shift, but, like all label information, should be truthful and not misleading about the nature of the ingredient. Proper strain curation, including deposit in biological resource centers, helps maintain the connection among the isolates tested and the ingredient in the food, and to maintain the properties and identity of the strain.
  • 3The panelists and speakers all seemed comfortable with the concept that it is necessary to identify and substantiate a health benefit specific to the strain specified on a label before the term probiotic can be properly used, although some members of the audience voiced discomfort with this as a regulatory requirement.
  • 4Considering that probiotics stand to benefit people in the management of a variety of health conditions, and assuming with such uses consumer safety is not a concern, it would be beneficial to have further discussions with the science, medical, consumer, commercial, and regulatory communities on how such benefits can be identified, substantiated, and effectively communicated.
  • 5Confusion among stakeholders led them to request that the FDA consider providing further guidance to industry on
    • a. approaches to the substantiation of structure/function claims specific to probiotics, such as in the target areas of digestive health and immune function; and
    • b. study designs that are physiologically meaningful, address all meanings of the claim, but do not trigger drug status.
  • 6Further discussions and guidance, as outlined above, might stimulate additional research in the field of probiotics, which in turn would provide consumers and healthcare professionals with information they seek to help them make informed choices about probiotics and their health.

Acknowledgments

This conference report was made possible, in part, by support from the National Center for Complementary and Alternative Medicine, a component of the National Institutes of Health. The views expressed in written conference materials or publications and by speakers and moderators do not necessarily reflect the official policies of the National Institutes of Health; nor do mention of trade names, commercial practices, or organizations imply endorsement by the U.S. government.

Conflicts of interest

MES consults with numerous food and dietary supplement companies conducting business in the probiotic industry. She does not have any ownership role or serve on governing boards for any company.

Ancillary