Why it is important to have a DRI-like process for the evaluation of bioactives

Bioactives are important to human health, they are an active area of research, and consumers are purposefully purchasing foods containing them. Substantial evidence exists that specific bioactives beneficially affect health. This conference heard from three experts on bioactives: Dr. Cesar G. Fraga on flavanols; Dr. Joseph Levy on lycopene and other tomato carotenoids; and Dr. Mark Messina on soybean isoflavones. A few of their major points are discussed here, since the overall goal of the conference was to discuss the feasibility of applying a dietary reference intake (DRI)-like process to the evaluation of bioactives rather than a scientific discussion on bioactives themselves, the reader is referred to a number of key papers for more information in support of a specific bioactive and decreased risk of a disease or other health-related condition. There is strong evidence for the effect of flavanols on decreased risk of cardiovascular disease and associated risk factors. This evidence includes demographic data and human interventions, and it is mechanistically supported by animal and ex vivo studies [10]. (-)-Epicatechin is the compound better studied [24]; however, other flavanols and flavonoids could share these protective actions [30]. Another important class of bioactives is isoflavones derived from soybean. Although it is not possible to infer a direct causal relationship, case–control and prospective epidemiologic studies show isoflavone intake via soyfoods is associated with a lower risk of several chronic diseases including breast [32] and prostate [33] cancer, and among women, coronary heart disease (CHD) [18] and osteoporosis [17, 34]. Furthermore, there is relatively solid evidence that isoflavones increase flow-mediated dilation in post-menopausal women with impaired endothelial function [6] and there is suggestive, but limited evidence, that isoflavones reduce carotid intima media thickness [12]. The most impressive clinical data exist for the alleviation of menopausal hot flashes [27]. Lycopene and other tomato carotenoids have been found to decrease blood pressure in pre-hypertensive patients as well as reduce post-prandial blood-oxidized low-density lipoproteins [5, 7, 16, 22].

Research on bioactives is a significant portion of diet, nutrition and disease portfolios of governments, at universities, and at food manufacturers. Consumers are interested in optimal health and are purposefully purchasing foods containing bioactives. However, there is no evaluative process in place to inform the public about the strength of the science behind the purported benefits of a specific bioactive of interest, nor is there information on how much of a particular bioactive is necessary to be of benefit. If there were a process to evaluate the strength of the science behind the intake of a bioactive and decreased risk of disease (or other health condition), standards would be set for this research, studies could be compared across laboratories, and consumers and health professionals could have more confidence in what they were eating; and the field could move forward more quickly. If that science base were combined with a recommended intake value, assessments could be made as to whether or not populations or specific age groups were meeting that recommendation and consumers would know the overall contribution of a food product to the recommended amount.

Having a DRI value increases the status of a bioactive and makes it part of nutrition public policy. Without a DRI value, it is unlikely that bioactive information will be incorporated into national nutrition intake surveys such as NHANES (National Health and Nutrition Examination Survey) in the US. National nutrition surveys describe the amount of nutrients being consumed by representative populations, and then those intake values are compared to a DRI to determine whether the population is eating too much or too little of that substance. If too little, the substance might be called an “at risk” nutrient, and education campaigns to improve people’s intakes (within one’s calorie allotment) could be implemented. Thus, not having reference intake values limits the ability to develop messaging to the public regarding bioactives for which there is solid scientific evidence of their health-enhancing effects. Importantly, health professionals (such as physicians, physician assistants, nurses, and dietitians) who may offer advice to clients on what they should be eating, would be more comfortable recommending bioactives if they have gone through a rigorous evaluation process. In most countries, the overall nutrition policy is called “dietary guidance”. Although this guidance is food rather than nutrient based, the food recommendations are derived from the DRI values for the nutrients. For example, the philosophy of the US Dietary Guidelines is that if one follows the recommendations of the guidelines, one will automatically meet the DRI values for all nutrients [20]. Thus, dietary guidance is another important way that information on bioactives with substantial science behind their efficacy could be transmitted to consumers. For a summary of the advantages of having a DRI-like process for the evaluation of bioactives, see Table 1.

Table 1 Why it is important to have a DRI-like process for the evaluation of bioactives

Dietary fiber is an example of a bioactive with a DRI value. Although dietary fiber is a non-essential nutrient, it does have an officially recommended intake value [14]. This means that the amount of fiber in a food product is on most fact-based food labels throughout the world. It is also generally included in the questionnaires on national food intake surveys so that information is available as to whether or not the DRI value for fiber is being met. It also means that it is considered and promoted in dietary guidance. Dietary fiber is thus of concern to consumers who are looking to increase it in their diets.

What are the challenges to establishing such a system and how can those challenges be met

The process for determining nutrient reference values in the US and Canada changed significantly in 1994 when several kinds of reference values were introduced and articulated in the 1994 publication, How Should The Recommended Dietary Allowances Be Revised [9]. There were two major changes: (1) that values could be based on reduced risk of a disease and (2) that there were additional values other than the recommended dietary allowance (RDA), i.e., estimated average requirement, adequate intake (AI), and upper level (UL). A conclusion of this report was that the “reduction in risk of chronic disease is a concept that should be included in the formulation of future RDAs where sufficient data for efficacy and safety exist [9] ”. This conclusion represented a “new paradigm” from what had previously existed. Using these criteria, four DRI values have since been set based on chronic disease: osteoporosis and fractures for calcium and vitamin D, dental caries for fluoride, CHD for fiber, and a combination of endpoints including salt sensitivity, kidney stones, and blood pressure for potassium [28]. Thus, this suggests that bioactives could qualify for a DRI value if they could show strong science behind reduced risk of disease.

Demonstrating reduced risk of disease with a bioactive is more difficult than it is to show prevention of a deficiency outcome with an essential nutrient. A major difference between bioactives and essential nutrients (i.e., vitamins, minerals, essential fatty acids, and essential amino acids) is that the absence of bioactives in the diet does not result in a deficiency disease, whereas the absence of an essential nutrient eventually results in deficiency symptoms (e.g., lack of vitamin C and scurvy, thiamin and beriberi, iron and anemia). This difference means that a DRI value would have to be based on an endpoint other than a deficiency disease. As shown above, this could be decreased risk of a chronic disease, but showing cause and effect with a bioactive and chronic disease is more difficult than when the disease is specific nutrient related. In other words, if vitamin C intake is inadequate, 100 % of the deficient people will eventually get scurvy. This is not the case for chronic disease which is affected by multiple nutrients, and is also impacted by other non-nutrient factors (e.g., gender, age, and genetics) [28].

Dr. Ben van Ommen challenged the concept of relating health to just decreased risk of disease and suggested that in quantifying the health effects of bioactives “we might need to consider in greater depth what health is, what mechanisms are involved in maintaining health, and how to best quantify these”. A pioneer in this new area of “optimal health,” he considers health to be appropriate adaptation to a continuously changing environment—and food is a key part of that changing environment. He calls this adaptive capacity “phenotypic flexibility” and states that it is key to maintenance of overall homeostasis “and thus to a healthy life”. He and his research group have also developed ways to test for “phenotypic flexibility” by stressing specific components of the system that maintain homeostasis and evaluating the stress response reactions. These response reactions usually appear to be more informative and sensitive than their homeostatic counterpart. A classic example is the oral glucose tolerance test versus fasting glucose, and numerous other comparable “challenge biomarkers” that are now being developed [23, 29]. If accepting decreased risk of disease as an endpoint for a DRI value was a paradigm shift, Dr. van Ommen’s emphasis on “phenotypic flexibility” is definitely a new paradigm shift which should become more widely accepted as an evaluation of efficacy for a bioactive as the research to measure this flexibility is validated.

Issues regarding the setting of life-stage DRI values for bioactives

Dr. Stephanie Atkinson discussed possible approaches for determining life-stage DRI values for bioactives using information from previous DRI recommendations developed for infants, children, and youth as an example [1]. She suggested using three age groupings to establish DRI values for bioactives: (1) infants to 1 year of age; (2) children 1–8 years; and (3) individuals over 8 years. For Infants to 1 year of age, she suggested using human milk as a “reference”. For children 1–8 years, in the absence of clinical trials, it was suggested that AI values be derived from population-based intake data associated with health outcomes. For those over age eight, the suggestion was to derive the value from existing data on biomarkers of chronic disease or extrapolation from adults. The rationale and the cautions for each of these recommendations were provided. Issues in establishing life-stage DRIs for bioactives vary greatly from one substance to another. For example, in infants, intakes from human milk bioactive substances such as nucleotides [25], carnitine [2], lutein [3], and glycoconjugate sugars [19, 26] have been used to derive safe levels of addition of such compounds to infant formulas. Evidence of the biological benefit of addition of these substances to the health of formula-fed infants is inconsistent, but no adverse effects have been identified. A lack of response to addition of a bioactive to formula may relate to the variable bioavailability of a bioactive depending on whether it is found in breast milk or added to formula. For example, approximately four times more lutein is needed in infant formula than is naturally present in human milk to achieve similar infant serum lutein concentrations [3]. For the case of dietary fiber, using human milk as a reference for infants 7–12 months of age cannot be done because of the absence of this substance in milk. Remaining challenges include selection of the best model (approach): For example, objectively differentiating between the various age groups on a basis other than age itself seems logical, if difficult. Also, the development of recommended intakes or maximal effect ranges is another choice.

Establishing safety of bioactives and adjusting for different population groups may not be the same as it is for essential nutrients

Dr. David Richardson discussed the process of establishing the safety of bioactives. For nutrients and other dietary ingredients, the limitations on safety are commonly set through identifying a “Tolerable Safe UL”. This is done by identifying any “hazard” associated with high intakes, establishing a dose–response relationship, evaluating the uncertainty and selecting a composite “safety factor,” and then calculating an UL value. This procedure cannot be applied when no hazard can be identified (as with many bioactives). However, there is an alternative risk assessment approach that is based on the highest observed intake (HOI) method developed by FAO/WHO [8] and included in Codex Guidelines [6]. The HOI is defined as the highest level of intake observed with the available data of acceptable quality, showing an absence of adverse effects. Since most bioactives have no known hazard, the HOI is an important alternative approach to setting quantitative value limits on the amounts of bioactives that may be considered safe.

Even if there is agreement on this general approach, the problem remains that most safety data are derived from studies on adult subjects designed to look for benefit rather than harm. Scaling the healthy adult values to give confident estimates of the amounts to be deemed safe in subpopulation groups is difficult. Nonetheless, an adult UL or HOI value is needed to give an appropriate basis for policies directed to other population groups. Most DRI values fall well below the ULs/safe ULs, but some high intakes can approach or exceed the safe UL. A narrow range between a DRI and upper safe level may be unjustified when there is a lack of evidence of a demonstrable adverse effect/toxicity at current levels above an upper safe level. If intakes exceed the UL/HOI, the significant uncertainties about the safe level are more likely to indicate that the intake is not the problem but rather the application of a safe level based on inadequate data. In practical terms, adverse effects are more often observed with inadequate intakes rather than excessive intakes. Clearly, care and scientific judgment must be taken in the use of a safe UL as the benchmark in the selection of ULs/HOIs for bioactives.

A sustainable approach is needed for the evaluation of efficacy and intake recommendations for bioactives

Lessons learned from South Korea

There is growing interest in establishing a DRI-like system for setting intake values for bioactives [4, 11]. Although South Korea does not have a DRI system for establishing intake values for bioactives, they do have an evaluative process together with a process to determine intake values. Dr. Namsoo Chang explained this process for South Korea. The Health Functional Food (HFF) Act was enacted in 2004 with the goal of ensuring the safety of HFF with certain health claims for consumer information. At its inception the HFF covered products in the form of tablets, capsules, powders, granules, pastes, gels, jellies, and bars that were intended to enhance and preserve human health and contained one or more functional ingredients or constituents. In 2008, the scope was extended to include conventional foods and other diet supplements.

What is unique about the HFF act in South Korea is that unlike other countries, the government of South Korea is endorsing a particular product with a HFF “seal”. There are two types of HFF, generic and product-specific. The generic type (shown in Table 2) contains both 28 essential nutrients and 55 non-nutrients. Both the nutrients and non-nutrients are considered to have substantial efficacy and safety data to have been considered for the generic category. All of these substances listed on the generic health/Functional Food Code include health claims and intake recommendations. This generic type HFF is most analogous to establishing a process for evaluation of efficacy and intake values for DRIs, although they are not called DRIs by the South Korean Government.

Table 2 Functional ingredients listed in the South Korean Health/Functional Food Code (Generic Type)

If the bioactive is not on the generic type list of functional ingredients, then it needs to follow a process and receive approval. Manufacturers submit a dossier for comprehensive scientific evaluation of safety and efficacy, which is reviewed by the Government and Advisory Committees. The application must consist of any data on the history of safe use, manufacturing processes, recommend intake levels, toxicological data, clinical data, nutritional evaluation data, and bioavailability data.

Soy isoflavones (discussed in this conference) have a generic health claim which is that they help to maintain bone health. The isoflavone content of common soybean products in Korea is known, as is the isoflavone intake in South Korea. Although no safety data were available in Korea, the safe intake level for isoflavones was adopted from the Japanese standards. A recommended intake is set at 24–27 mg/day as aglycone soybean isoflavones. Notably, a caution is stated for infants, children, pregnant and lactating women, and individuals who have an allergy to soybean, and individuals who are sensitive to estrogen. A generic claim for lutein (a bioactive found in tomatoes) also exists. The health claim is, “helps eye health by maintaining the density of macular pigments which can be decreased by aging”. Based on review of existing literature, the intake recommendation was set at 10–20 mg lutein/day with a warning for yellowing of skin if taken at excessive amounts. In addition, a recommendation for intake of all-trans lycopene at 5.7–15 mg/day is provided based on the health claim for tomato extracts as an antioxidant. This recommendation is accompanied by a caution for pregnant and lactating women and for children. Flavonoids and lycopene are listed in the product-specific category, rather than the generic category. A flavonoid database is available for commonly consumed food by Koreans based on the USDA and Japanese flavonoid databases, which were developed in 2009. Flavonoids have been linked to reduced risk for chronic diseases and improved health outcomes, and six subclasses of flavonoids are identified by structure.

Currently, the Ministry of Health and Welfare is revising the South Korean DRIs and plans to release the revised version in 2015. Although it was recently decided that bioactive substances will not be included in the 2015 version of the DRIs, the need to establish DRIs for bioactive substances was raised. If there were to be a DRI value for bioactives, it would most likely be the AI value. The AI is defined as “The recommended average daily intake level based on observed or experimentally determined approximations or estimates of nutrient intake by a group (or groups) of apparently healthy people that are assumed to be adequate—used when an RDA cannot be determined” [13]. Importantly, South Korea may be able to contribute to establishing ULs for bioactives, since they have a post-market surveillance system on health/functional foods. They are operating an online system for adverse events data collection from consumers, manufacturers, and healthcare professionals. They have the integrated database on products and safety data and are in the process of doing statistical modeling to determine a cause effect relationship of any adverse event.

Setting specific proposed levels for bioactive compounds: Recent experiences in China

Professor Yang Yuexin described the process for setting a special category of DRIs (called specific proposed level; SPL) in China. This new category is used to evaluate and assign an intake value for bioactives. This is the only country, of which we are aware, that has actually established DRI values for bioactives. The China Nutrition Society, similar to the Institute of Medicine in the US, changed their intake evaluation process for nutrients from only RDAs to DRIs. This change was initiated in 2000 and resulted in 32 DRI values for nutrients. In 2010, they initiated the incorporation of a SPL for non-nutrients and a proposed Intake that is based on reducing the risk of non-communicable chronic disease and improving optimal health. Their stated rationale as to why they consider the SPL a DRI value is that both traditional medicine and modern nutrition research have deepened the understanding of plant compounds; and also because consumers are widely consuming these bioactive substances in China. In 2010, they had seven different expert review panels containing a total of 87 experts develope the DRIs for China to be released in 2014. One of the seven panels was on “non-nutrients,” and 21 experts were involved in this panel. The goal of this panel was to develop DRI values for water, fiber, and 18 phytochemicals (SPLs). The SPLs reflect the current state of scientific knowledge and are published as a series of reports by the Chinese Nutrition Society. Both SPLs and ULs are set for bioactives. Table 3 shows the “non-nutrients” that were evaluated by the Chinese DRI process.

Table 3 “Non-nutrients” that were evaluated by the Chinese DRI Process

The Chinese Nutrition Society has acknowledged that there are some bioactives that “like some other nutrients, are essential for reaching the full (genetically-determined) lifespan”. They have termed these nutrients as “life span essential” [31]. The Chinese experience in establishing DRI-like values for bioactives should be followed closely, and they should be acknowledged as being the pioneers in this area.

Setting a high bar for entrance into the evaluation system

One issue with setting up a DRI-like process for the evaluation of bioactives is the very wide range of the strength of the science behind the intake of a bioactive and a purported reduced risk of disease. For some bioactives, little research has been conducted, whereas for others there are 20–30 years of research in support of a protective effect. A concern is that the evaluators would have to be dealing with requests when there was insufficient information to apply the process. One suggestion to offset this challenge is to set a high standard for “entrance into the evaluative process”. Dr. Joanne Lupton discussed potential entrance criteria as necessary information before a bioactive could be considered for a DRI-like evaluation process (see Table 4). Setting these nine criteria as essential for consideration for evaluation serves several goals: It minimizes the effort of the evaluator; and importantly, it sets a standard, if met, that investigators and funding sources could design their research to meet, knowing that there would be a certain level of credibility if they were to do so.

Table 4 Proposed criteria for a bioactive to qualify for evaluation

Summary, conclusion, and next steps

The speakers were in consensus that providing a framework for the evaluation of bioactives could be of benefit to scientists working in this field, to funders of the research, to governments, and importantly to consumers. However, they were also aware of the potential challenges to establishing such a framework. Clearly, there is a difference between determining intake values for essential nutrients and bioactives, and thus the basis of the intake value cannot be on a single-nutrient deficiency disease. Nonetheless, other endpoints such as reduced risk of disease may be applicable. Basing a DRI value on reduced risk of disease has been used for four nutrients that have DRI values. Alternatively, the AI value was considered by some to be an appropriate value for consideration as by definition it can reflect the current intake of specific healthy populations. Setting life-stage values for bioactives is also a challenge, but Dr. Atkinson suggested a different model for consideration. Instead of concentrating on the bioactive, per se, she suggested establishing goals for life stages. For example, for early life, it might be “optimal development” and markers for that could be body composition, or cognitive/behavioral outcomes. For child/adolescent, the DRI value could be based on early biomarkers that are sensitive indicators of chronic disease risk. Then, bioactives that were shown to affect those outcomes could receive intake values for that life stage. This model warrants development and consideration. Another challenge is establishing an UL value for bioactives in the absence of any evidence of toxicological effects. Here, it appears that there is an extensive literature on risk/benefit systems which should be considered for application to bioactives. Finally, the logistics of how to set the framework, who is the “keeper” of the system, and what it would take for a bioactive to be considered in this framework requires serious consideration. A proposed next step would be a workshop with representation from all key stakeholders to discuss the challenges to having a framework for the evaluation of bioactives and how those challenges may be overcome.