5
Looking to the Future Process for DRI Development: Session 41

Speakers in the final session were asked to speculate on the future process of Dietary Reference Intake (DRI) development. Specific topics of interest included enhancing transparency of the decision-making process, criteria and “triggers” for updating and reviewing DRIs, determining “new” nutrient substances for DRI development, options for stakeholder input, and important issues that may emerge in the future.

The moderator for the session was Dr. Paul Coates of the National Institutes of Health (NIH). Dr. Catherine Woteki of Mars Inc. opened the session with a presentation on emerging issues and future directions. Dr. Robert Russell of Tufts University then addressed the need to enhance the transparency of the decision-making process. Dr. Linda Meyers of the Institute of Medicine (IOM) gave a brief overview of the options for stakeholder input into the DRI process. Dr. John Suttie of the University of Wisconsin then outlined some types of criteria that could be used to determine when to update or review existing DRIs. Dr. Peter Greenwald of the National Cancer Institute (NCI) examined some considerations in specifying “new” nutrient substances for DRI study.

Discussions open to all audience members were held after each presentation, and a panel discussion in which panel members reflected on what they had heard about DRI development during the workshop closed the session.

1

This chapter is an edited version of remarks presented by Drs. Woteki, Russell, Meyers, Suttie, and Greenwald at the workshop. Discussions are composites of input from various panel members, discussants, presenters, moderators, and audience members.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 123
5 Looking to the Future Process for DRI Development: Session 41 Speakers in the final session were asked to speculate on the future process of Dietary Reference Intake (DRI) development. Specific topics of interest included enhancing transparency of the decision-making pro- cess, criteria and “triggers” for updating and reviewing DRIs, determining “new” nutrient substances for DRI development, options for stakeholder input, and important issues that may emerge in the future. The moderator for the session was Dr. Paul Coates of the National Institutes of Health (NIH). Dr. Catherine Woteki of Mars Inc. opened the session with a presentation on emerging issues and future directions. Dr. Robert Russell of Tufts University then addressed the need to enhance the transparency of the decision-making process. Dr. Linda Meyers of the Insti- tute of Medicine (IOM) gave a brief overview of the options for stakeholder input into the DRI process. Dr. John Suttie of the University of Wisconsin then outlined some types of criteria that could be used to determine when to update or review existing DRIs. Dr. Peter Greenwald of the National Cancer Institute (NCI) examined some considerations in specifying “new” nutrient substances for DRI study. Discussions open to all audience members were held after each presen- tation, and a panel discussion in which panel members reflected on what they had heard about DRI development during the workshop closed the session. 1 This chapter is an edited version of remarks presented by Drs. Woteki, Russell, Meyers, Suttie, and Greenwald at the workshop. Discussions are composites of input from various panel members, discussants, presenters, moderators, and audience members. 12

OCR for page 123
124 THE DEVELOPMENT OF DRIs 1994–2004 EMERGING ISSUES: WHAT NEW CHALLENGES MIGHT THE FUTURE HOLD? Presenter: Catherine Woteki DRI development is an important scientific undertaking. Future activi- ties can be informed by the conceptual evolution and changes in science that have occurred over nearly 100 years of establishing dietary recom- mendations for populations. Moreover, it is important to recognize that there is a long tradition of providing guidance on a health promoting diet. Scientists as early as the late 1700s appear to have included concerns about health in advice to the U.S. Congress concerning the creation of the Navy Ration Law. Later, in the 1800s, the U.S. Army Surgeon proposed limit- ing fat intake and eating a balanced diet; at about the same time the U.S. Department of Agriculture (USDA) highlighted the use of a balanced diet to promote health. Emerging Issues: 1925–1990 Government recommendations for what constitutes a health-promoting diet can be readily traced to the World War I era, when the British Royal Society produced recommendations for food requirements for populations under stress during wartime. During the period 1925–1937, the Health Organization of the League of Nations began to set estimated requirements for specific vitamins and minerals. In 1933, dietary standards for food programs were proposed by both the British Medical Association and the USDA. An enormous conceptual evolution occurred from 1917 to 1937, with dietary recommendations changing from a focus on food for starva- tion relief to standards for programs to maintain and improve the health of the population. A second conceptual evolution was the move from basing recommenda- tions on observations of usual food intake to using the emerging scientific knowledge of the needs for essential nutrients and energy to provide specific nutrient recommendations. From 1941—when the Roosevelt administra- tion asked the National Research Council (NRC) to recommend levels of intake of essential nutrients to maintain the health of the population dur- ing wartime—until 1989, 10 editions of the values (called Recommended Dietary Allowances [RDAs]) were issued. Although the focus was on essential nutrients and establishing estimates of intake to maintain health, concepts of disease prevention were addressed in editions of the RDAs as early as 1958 (NRC, 1958) when a relationship between dietary fat and coronary artery disease mortality was noted. The 1964 edition (NRC, 1964) recommended that adults moderately reduce

OCR for page 123
12 LOOKING TO THE FUTURE PROCESS FOR DRI DEVELOPMENT total fat intake and substitute polyunsaturated fatty acids for saturated fats. The 1968 edition (NRC, 1968) advocated for higher levels of physical ac- tivity to reduce the risk of arterial disease, obesity, and diabetes. The 1974 and 1980 editions (NRC, 1974, 1980) contained specific recommendations to decrease calories from fat to less than 35 percent, to decrease saturated fat to less than 10 percent, and to increase polyunsaturated fat intake. The 1989 edition (NRC, 1989a) recommended that the public and profession- als look to the IOM report Diet and Health: Implications for Reducing Chronic Disease Risk (NRC, 1989b) for more information about dietary intakes as they relate to chronic disease prevention. Emerging Issues: 1990–2004 As the IOM was embarking on the DRI process in 1994, several issues had emerged that framed the thinking. One was chronic disease risk reduc- tion. Several influential reports had been published, including Diet and Health: Implications for Reducing Chronic Disease Risk (NRC, 1989b) and The Surgeon General’s Report on Nutrition and Health (Office of the Surgeon General, 1988), and the idea that nutrition can play a role in reducing the risk of chronic disease was increasingly recognized. Interest grew in providing a quantitative basis for dietary guidance. A second issue was the concept of a safe range of intake. There was growing concern about voluntary fortification in the food supply and the increasing use of dietary supplements. It was recognized that there would be considerable value in establishing upper levels of intake and therefore, in a sense, establishing a range of intake conducive to good health. This issue had precedence. Some discussion about nutrient toxicity concerns can be found in the text of early recommended intakes from the NRC, dating back to the 1950s. In the late 1960s, the NRC together with the Council on Foods of the American Medical Association published a policy statement that included the need to set food fortification limits and supplement levels below harmful levels. New approaches also were considered. The United Kingdom’s Commit- tee on Medical Aspects of Food Policy had established multiple reference points in its report on dietary reference values. Dr. Beaton and others had done pioneering work on developing the probability approach. As the DRIs were being developed, new methods for informing public health policy deci- sions emerged. They included systematic evidence-based reviews (SEBRs), which arose in clinical practice to address the types of evidence appropriate for patient care recommendations; and quantitative risk assessment, which had emerged earlier to assess the risk of non-nutrient substances. Also during that period, results from clinical trials using nutrient interventions

OCR for page 123
12 THE DEVELOPMENT OF DRIs 1994–2004 against chronic disease endpoints suggested that the single nutrient/chronic disease prevention paradigm had largely failed. The Future We can anticipate considerable change and new approaches for mak- ing public health decisions in general, and specifically for developing DRIs. Some factors important to the DRIs of the future are described briefly below. Scientific progress will further our approaches, methodologies, and insights: • Undoubtedly scientific progress will move us from the consideration of single nutrients to patterns of nutrients or food intake, especially for chronic disease endpoints. • Divorcing chronic disease endpoint considerations from the estab- lishment of future DRIs will not be possible. The question is where it is appropriate to do so and how. • Scientific progress will also lead us to focus on “new” nutrient substances and on decision making concerning the adequacy of the science base for the purposes of DRI development for these emerging substances. • Better statistical techniques for extrapolation and scaling will be developed. • There will be greater insights with respect to physiological, environ- mental, and genetic factors. The public health context for setting DRIs will shift: • Chronic disease prevention was the overwhelming concern, along with nutrient excess, that framed the thinking in the DRI process. Now obesity is the public health concern through which all of our nutrition problems are being viewed. • Questions about appropriate levels of fortification and supplemen- tation and the role the DRIs play in informing those decisions will continue, as will questions about formulation of special foods for specific age groups. A major issue for the future will be the DRI development process: • Process problems that surfaced during the past 10 years include maintaining consistency across study committees as well as the time- liness, transparency, and openness of the process. In the future, revi-

OCR for page 123
12 LOOKING TO THE FUTURE PROCESS FOR DRI DEVELOPMENT sions organized by specific nutrients or groups of nutrients should allow greater transparency in addition to better risk characteriza- tion, consideration of uses, and verification of the “reasonableness” of those estimates as recommendations. • Guidance is also needed on how to set the DRIs, the role for SEBRs, statistical techniques, or other new methods to maintain consistency and scientific integrity. • The risk analysis paradigm provides an excellent model for a future DRI development process. However, it puts additional responsibili- ties on the government sponsors to provide a clear articulation of the uses of the DRIs. It also puts more responsibility on the IOM in terms of communication within the risk analysis paradigm, com- munication with the sponsors, and communication with the scientific community. Implications The issues raised about DRI development offer direction for the kinds of information needed and identify the steps to be taken to improve future DRIs. We have now identified the key issues, and we have the demonstrated need. Clearly, we can move forward, and clearly, the sponsors and the IOM must embrace larger responsibilities. Furthermore, it is hoped that govern- ment agencies that conduct research on nutrition will look closely at these recommendations and develop a concerted research program to address the gaps. The process itself can be improved, but better data are needed. Open Discussion An audience member said she shared Dr. Woteki’s hopes about the re- search agenda, but was not sure how it could be organized and who would lead. Dr. Woteki responded by pointing to the need for a broad partnership. She suggested that the starting point would be a rough prioritization of critical research needs and indicated that professional societies can play a key role in this activity. Just as importantly, the political will must be found to ensure funding and related support. In short, the approach must be a concerted effort from a scientific and health perspective. Dr. Woteki agreed with one commenter who expressed the opinion that the burgeoning interest in functional foods—and, in turn, claims about these ingredients that communicate a sense of requirement—suggest a need for some parallelism between these emerging issues and the DRI develop- ment process. Another audience member noted a recurring theme concerning the need for government sponsors of the DRI process to articulate their uses for

OCR for page 123
12 THE DEVELOPMENT OF DRIs 1994–2004 DRIs. She expressed the opinion that this could lead to policy influencing the science and suggested that the DRI study committees should operate independently from issues surrounding uses of DRIs. Dr. Woteki explained that the risk analysis framework she described allowed the independent and unbiased review of the science, but that it also articulated a clear role for sponsors of the assessment to specify the nature of the problem that the assessment is intended to address. Unless those reviewing the science understand why they have been requested to carry out their activities, their outcomes will not be useful. She pointed out that government representa- tives have expressed concern in the past about the unresponsiveness of the DRI outcomes to their needs and are increasingly recognizing that this may be a function of assigning relatively vague tasks to study committees. Dr. Woteki suggested that the decisions about the nature of questions to be directed to study committees are an enormous and complex responsibility that, under the risk analysis paradigm, includes input from sponsors of the DRI process. IS THERE A NEED TO ENHANCE TRANSPARENCY OF THE DECISION-MAKING PROCESS? Presenter: Robert M. Russell The interest in enhancing the transparency of the scientific decision- making process is by no means unique to the development of DRI values. Others have noted that, as a general matter, a lack of transparency can result in • perceived inconsistency (whether or not it exists); • perceived lack of objectivity (e.g., prejudices of study committee members); • complexity in presentation (rambling narratives that do not provide much information); • a lack of clarity; • difficulty in implementation; • decreased chances of replicability (from one study committee to another); and • hidden research gaps (Garza and Pelletier, 2007). Our 10 years of experience indicate there are several points in the DRI process where efforts to make the decisions more transparent would have helped with the clarity of the outcomes and with more ready acceptance of the inevitable scientific judgments needed. Moreover, more information about the decision-making process would have mitigated concerns that

OCR for page 123
129 LOOKING TO THE FUTURE PROCESS FOR DRI DEVELOPMENT study committees did not consider or review certain data or options. In fact, these may have been well considered, but there was no related discussion in the text or “transparency trail” to make this known. Examples of Transparency Issues Criteria for Literature Inclusion/Exclusion One transparency concern that requires attention is the lack of com- plete “up-front” documentation on the criteria used for literature searches (e.g., what literature was excluded or included). The volume of literature searched was often not explained well in the narratives. Criteria for Evaluating Evidence The criteria for evaluating and weighing evidence were also not well documented. Three brief examples—vitamin B6, zinc, and β-carotene— illustrate this issue. Panels in Britain, North America, and the European Union (EU), looking at the same databases and using the same framework, came up with quite different conclusions. The reasons for this are unclear because the narratives in their reports are not explicit. For vitamin B6, in North America, the tolerable upper intake level (UL) was based on a study by Bernstein and Lobitz (1988) in which people taking large amounts of pyridoxine exhibited neurological abnormalities at over 200 mg/day. A no-observed-adverse-effect level (NOAEL) of 200 mg/day was chosen and divided by an uncertainty factor of 2 based on a small number of data on doses less than 200 mg/day. The UL was therefore 100 mg/day. In Britain, they used a study on dogs (Philips et al., 1978) in which the dogs developed ataxia at 3,000 mg/day; that was taken as a lowest-observed-adverse-effect level (LOAEL) and divided by an uncertainty factor of 300 (because of the LOAEL and interspecies and interindividual variation) to derive a UL of 10 mg/day. The EU UL was based on a study by Dalton and Dalton (1987) on self-reported neurological symptoms; the LOAEL of 100 mg/day was divided by an uncertainty factor of 4, based on deficiencies of the database, to give a UL of 25 mg/day. Therefore, ULs of 100, 10, and 25 mg/day were derived by three panels looking at the same data. For zinc, in North America, the UL was based on a study by Yadrick et al. (1989), which showed a decrease in erythrocyte superoxide dismutase at 60 mg of total intake (diet plus supplementary), to derive a UL of 40 mg/day using an uncertainty factor of 1.5 to account for interindividual variation and use of a LOAEL. In Britain, using the same study, they came up with a LOAEL of 50 mg instead of 60 mg and derived a UL of 25 mg/

OCR for page 123
10 THE DEVELOPMENT OF DRIs 1994–2004 day using an uncertainty factor of 2 for use of a LOAEL. The EU rejected that study and instead used balance studies (Davis et al., 2000; Milne et al., 2001), deriving a UL of 25 mg/day. For β-carotene, in North America, no UL was established because there was no dose–response and the study committee believed the data were conflicting (with no clear explanation). In Britain, they based the UL of 7 mg/day on the Alpha-Tocopherol Beta-Carotene (ATBC) study (and backed this number up by a study in ferrets) in which the equivalent LOAEL was 20 mg/day. In the EU, no UL was established because there were no dose–response data and because different formulations were used in the various studies. Another example is vitamin A, for which the Food and Agriculture Organization of the United Nations/World Health Organization Recom- mended Nutrient Intake differs from the IOM RDA (500 µg versus 700 µg for 19- to 50-year-old females). The narratives do not explain why the recommendations differ. These problems may be solved if SEBRs are used to assist in deriving the endpoints for Estimated Average Requirements (EARs) and ULs. The reviews document the scientific evidence or the evaluation, document and rank the uncertainties around the estimations, and rank the health implica- tions of the intakes above and below the reference intake. There is provision of a rationale, in the form of extensive tables, for decisions reached, replac- ing the rambling and vague narratives in some reports. Subjective Uncertainty Factors The factors used to account for various sources of uncertainty were highly subjective and varied depending on the study committee. Bias can be minimized, however, by following predefined rules—not by using an uncertainty factor that aims for a convenient UL or a UL above the recom- mended intake, for example. Endpoint Selection More transparency would have helped with the concerns that surround the selection of endpoints used. Although there is a presumption that these choices were made for public health protection and significance, the deci- sion process is not clear from the narratives. Specification of Research Gaps An important activity carried out without benefit of an identifiable and accountable set of criteria was the specification of research gaps. Appar- ently the research gaps identified by study committees were often compiled

OCR for page 123
11 LOOKING TO THE FUTURE PROCESS FOR DRI DEVELOPMENT at the last minute, after the committees had completed their exhaustive work of deriving EARs and ULs. In short, research gaps were developed too rapidly and without enough deliberation. An overall effort at transpar- ency would have called attention to this approach and perhaps improved it, making it more deliberative. Configuration of Study Committees Finally, improved transparency has the potential to ameliorate the nearly inevitable problems associated with configuring study committees. The configuration of study committees may introduce bias, particularly when the committees (which were limited in size) had to deal with large numbers of nutrients. For example, the micronutrient committee had 14 members to study 14 nutrients. For a number of nutrients, there was only one expert on the committee. It is possible that a single person’s opinion could go unchecked, particularly if he or she had an assertive personality. Transparency and accountable documentation for the decision made could help in terms of double-checking outcomes and explaining the reasons behind decisions. Implications The DRI development process would be improved, and transparency enhanced, if more active and targeted efforts were made to explain and document the decisions made. More specifically, SEBRs should be used to help make key decisions on endpoint selection, for both EARs and ULs, with thorough documentation tables. Rating scales for uncertainty and public health importance should be used as an important part of the evi- dence-based review process. Also, predefined rules should be followed for uncertainty factors to minimize bias. Although transparency can help to mitigate the effect of one strong member of a study committee when it may have only one expert on a certain topic, a better solution might be to focus on a smaller subset of nutrients—a single nutrient or small groups of nutrients that interact (e.g., vitamin D, calcium, phosphorus, the antioxidant nutrients; folate, vitamin B12, vitamin B6, riboflavin; sodium, potassium)—so that more depth in expertise can be available on the committee and knowledgeable group discussions can be used to reach conclusions rather than the perspective of a single scientist. Open Discussion A participant expressed doubt that SEBRs would be helpful to DRI development. He questioned how scientists who are not nutritionists could

OCR for page 123
12 THE DEVELOPMENT OF DRIs 1994–2004 judge the relative importance of studies germane for setting reference val- ues. Dr. Russell replied that SEBRs could enhance some critical compo- nents of the development process, but do not replace activities that would focus on scientific judgment and decisions made by those with nutritional or other appropriate expertise. Furthermore, such reviews would not be needed—nor are they necessarily appropriate—for every decision, every nutrient, or every DRI value. Another participant suggested the drive for consensus in the DRI pro- cess may have obscured important disagreements. She asked whether en- hancing transparency, as described by Dr. Russell, would avoid this, and indicated it may be better to explicitly state and explain the disagreement rather than forge a consensus. Dr. Russell responded that an SEBR would have helped the vitamin D discussions, for example, with regard to the ap- propriateness of the 25-hydroxyvitamin D level as an indicator. He added that there is considerable value in making available an objective scientific evaluation of the data via an SEBR rather than depending on the potentially biased opinions of strong personalities. An audience member commented that the inability to openly discuss challenges is problematic and could limit transparency. It was suggested that there could be points in the process when outside advice should be sought. Dr. Russell agreed. In the future a feedback loop may be possible, as long as there is no possibility that stakeholders could influence the science or compromise the scientific independence of the study committee. One participant noted that, with respect to the β-carotene example highlighted by Dr. Russell, a clear problem formulation step or an indica- tion of the uses of the needed reference values, particularly regarding the intended population, would have helped to enhance clarity. She then asked about the composition and functioning of study committees, suggesting there were two models. In the first model, study committee members are likely to have been investigators for some of the studies that will be re- viewed and used in deriving the DRIs. In the second—a consensus model approach, as is common at institutions such as NIH—the study committee members have relevant expertise, but their studies will not be under review in terms of the scientific evaluation. She asked which would be more help- ful for transparency and scientific rigor. Dr. Russell responded that study committees should have a meaningful number of people who are specific experts, and expressed concern that a consensus-type approach would not solve all the problems associated with study committee bias, adding that more than two members on a study committee per nutrient topic area seemed to be another important avenue to pursue.

OCR for page 123
1 LOOKING TO THE FUTURE PROCESS FOR DRI DEVELOPMENT OPPORTUNITIES FOR STAKEHOLDER INPUT IN IOM ACTIVITIES Presenter: Linda D. Meyers This presentation focuses on the “rules” surrounding study report de- velopment at the IOM and the options for input. To address these topics, we need to understand what the IOM is, why advice is sought, how IOM functions, and what motivates the procedures for input. Role of the Institute of Medicine The IOM is generally regarded as a trusted independent advisor. The government approaches the IOM for consensus advice that is not influ- enced by any particular group (including the government) and that draws on the best minds in the country. The IOM was established in 1970 under the charter of the National Academy of Sciences. It is part of the National Academies complex that encompasses three honorific societies: National Academy of Sciences, National Academy of Engineering, and Institute of Medicine. The IOM is a private, independent, nonprofit, “soft money” organiza- tion, with no line appropriation from Congress. It has 1,600 elected mem- bers and about 130 staff in 6 program units, one of which is the Food and Nutrition Board (FNB). It publishes about 40 reports a year. The IOM—and the entire National Academies complex—engages in ac- tivities that include committee studies, workshops, forums and roundtables, symposiums and lectures, expert meetings, and communication functions. Communication activities include so-called “derivative products,” which are based on existing reports and contain no new recommendations. Nature of Consensus Report Development and Opportunities for Input Reports produced by consensus committees tackle major health issues, such as DRIs, obesity prevention, overhaul of the Food and Drug Admin- istration’s drug review system, medical errors, and health literacy. As the backbone of the Academies’ activities, the committee studies are the most visible. They are prepared by balanced expert committees. Experts serve without remuneration, an effort designed to increase their independence. The committees strive for full consensus based on evidence, and their work is evaluated through a rigorous peer review process. Many current policies for the process of report development derive from Section 15 of the Federal Advisory Committee Act (FACA). Section 15 was passed in 1997 and is intended to maintain the Academies’ indepen- dence from government. It requires the Academies to ensure public input

OCR for page 123
140 THE DEVELOPMENT OF DRIs 1994–2004 Randomized Controlled Double-Blind Randomized Studies Controlled Studies Cohort Studies Case-Control Studies Case Series Case Reports Ideas, Opinions FIGURE 5-2 Amount of research done, by type: The smallest amount of research 5-2.eps is conducted using randomized clinical trials (most useful). our purposes, it is important to note that the most definitive studies are randomized controlled clinical trials (buttressed by basic nutritional sci- ence), followed by non-randomized controlled trials. The studies become weaker from that point, with cohort or case-control studies and ecologic studies appearing as one moves down the pyramid. At the very bottom of the list are the opinions of respected authorities. Figure 5-2, developed by Dr. Barry Kramer, underscores the challenge we face. The pyramid shows the amount of research in each area mentioned above. Little research of the most useful type (randomized clinical trials) is available, whereas there is an enormous amount of information that is not very meaningful. This needs to be reversed. Nutrient Substances: The Nature of the Evidence As we learn more about individuals and individual variation in genom- ics and related fields, we find changing evidence for nutrient–disease rela- tionships. We have to develop an approach to consider the changes people experience over their lifetimes, their different sustainabilities and reactions

OCR for page 123
141 LOOKING TO THE FUTURE PROCESS FOR DRI DEVELOPMENT to foods, and the changing environments to which they are exposed. To illustrate these points, two studies will be discussed. Linxian Nutrition Intervention Trial The first study is a 5-year trial started 25 years ago in collaboration with Chinese scientists and conducted in rural China, where the diet is bor- derline deficient and the population is very stable. The main focus was on esophageal cancer, including cancers of the upper stomach. Relevant study information includes the following: • Study population: Nearly 30,000 adults aged 40–69. • Study design: Factorial for comparison of four groups of nutri- ents: (1) retinol and zinc; (2) riboflavin and niacin; (3) ascorbic acid and molybdenum; and (4) selenium, β-carotene, and vitamin E (factor D). • Outcomes: Factor D decreased total mortality by 9 percent (Fig- ure 5-3), decreased total cancer mortality by 13 percent, and de- creased total gastric cancer mortality by 21 percent (Blot et al., 1993). However, whereas factor D decreased total mortality for those who were under 55 at the start of the trial (1986), the benefit largely occurred after the intervention was stopped (1991). The fact that the time of exposure may not be the same as the time of the benefit must be considered with nutrients. In the people aged 55 and above, it made no difference at all. CDF CDF Log-rank P=0.001 Log-rank P=0.443 0.6 0.19 RR=0.88 RR=0.98 0.18 Placebo Placebo 0.17 0.16 0.5 0.15 0.14 0.13 0.4 0.12 Factor D 0.11 Factor D 0.10 0.3 0.09 0.08 0.07 0.2 0.06 0.05 0.04 0.1 0.03 0.02 <55 years 55+ years 0.01 0.00 0.0 1986 1988 1990 1992 1994 1996 1998 2000 2002 1986 1988 1990 1992 1994 1996 1998 2000 2002 FIGURE 5-3 Linxian Nutrition Intervention Trial: total mortality decreased in <55 5-3.eps years age group by factor D.

OCR for page 123
142 THE DEVELOPMENT OF DRIs 1994–2004 Also in this study, a similar pattern was observed for total cancer mor- tality and stomach cancer mortality (Taylor et al., 2005).3 This suggests that the intervention had an impact in the precancerous period. Complicating the issue further was the observation that what benefited overall mortality, total cancer mortality, and stomach cancer mortality appeared to be detri- mental against esophageal cancer mortality. This indicates that the overall effect must be considered if assessing the effects of nutrients on a popula- tion at large or major subgroups of the population. U.S.–Finland ATBC Lung Cancer Prevention Trial This was a study using 29,000 men who were heavy, long-term smokers. In the early 1980s, when the trial started, most epidemiologists believed that β-carotene prevented cancer. The trial was factorial. The participants took vitamin E, placebo for vitamin E, β-carotene, and placebo for β-carotene. There was no difference in lung cancer incidence for about 4 years. After that, however, the curves started to separate (Figure 5-4), and smokers on β-carotene began to do worse. In the β-carotene group, the risk of lung cancer was about 6 per 1,000 men per year, whereas in the placebo group, it was about 5 per 1,000 men per year, a 16 percent difference. The important message is that a large, well-designed, well-managed, double-blind clinical trial was needed to detect that 1 per 1,000 difference. Without this trial, people might still believe today that β-carotene protects against cancer. There was also a one-third reduction of prostate cancer occurrence in this trial in the men taking vitamin E. With about 14,500 men on placebo for vitamin E, 147 got clinical prostate cancer. With 14,500 on vitamin E, 99 got clinical prostate cancer. As further evidence of the value of this type of work, the observation of a secondary endpoint, together with other data, led to the design of a study of 32,000 men now in progress, to examine whether vitamin E and/or selenium will reduce the occurrence rate of clini- cal prostate cancer. Implications Regarding efforts to establish the impacts and risks/benefits of nutrient substances, the following points are worth noting: • Efficacy data based on sound scientific evidence must be present before making public health recommendations regarding nutrients. 3 Also personal communication, P. Taylor, National Cancer Institute.

OCR for page 123
14 LOOKING TO THE FUTURE PROCESS FOR DRI DEVELOPMENT FIGURE 5-4 Cumulative lung cancer incidence in U.S.–Finland ATBC Lung Cancer Prevention Trial shows that beta carotene does not protect against cancer. 5-4.eps bitmap image embedded in very dark background— inverted colors for legibility, but need a better original • Many existing data are not sufficient, not sound, and even contradic- tory; these need to be sorted through using systematic approaches. • Confidence in nutrient–disease relationships can change, often in unexpected directions. • Large randomized trials have the greatest impact in changing the level of confidence in a nutrient–disease relationship. Although these trials have an enormous cost, they are necessary. • We need greater investment in research in the nutrition area. To underscore the importance of “getting the science right,” we need only turn to a recent article in the New York Times Magazine written by a respected science reporter. It was entitled “Why can’t we trust much of what we hear about diet, health and behavior-related diseases?” (Taubes, 2007). The reporter includes several examples, many in nutrition epidemiology, where there is so much conflicting evidence that people do not believe it. Clearly, we have a serious problem, and we must push for the conduct of definitive studies before we make pronouncements on public health.

OCR for page 123
144 THE DEVELOPMENT OF DRIs 1994–2004 Open Discussion One commenter noted that the research pyramid shown by Dr. Greenwald did not contain animal or in vitro studies. Dr. Greenwald re- sponded that although they were not included, animal studies are important if they are relevant to humans. Another participant added that understand- ing the underlying mechanism is one of the enormous values that comes from animal studies. Dr. Greenwald agreed, provided again that the studies are relevant to humans. An audience member asked Dr. Greenwald to comment on obtain- ing funding for research within the nutrition community at large. Dr. Greenwald responded that existing groups within the nutrition field, includ- ing the central coordination at NIH, tend not to be aggressive in looking for resources, and this shortcoming needs to be remedied. A participant emphasized his concern that randomized trials for nutri- tion lacked intermediary variables that in turn impact the plausibility of the effect. He stressed the desirability of establishing plausibility as it relates to clinical trial design. Dr. Greenwald responded that this was a good point, and that while biomarkers can play an important role, too often measures cited as biomarkers are only reflective of associations. Unless a biomarker is demonstrated to be predictive of the outcome of interest, such measures are not sufficient. An audience member expressed the opinion that there were functional food factors that were effective in preventing cancer and asked about the desirability of informing the public through food labeling. Dr. Greenwald replied that the data are too sparse to support such activities. He empha- sized that a true disease endpoint must be identified, covering multiple categories of disease over a sufficient period of time and taking into account individual variability. A participant asked about an approach for food components such as macronutrients that cannot easily be studied in randomized clinical trials or similar types of research. Dr. Greenwald responded that the needed invest- ment in basic nutritional science should specifically include efforts to clarify the role of such substances in different cells, organ systems, and models. He further suggested that there may also be a role for clinical metabolic stud- ies to provide initial and clarifying data on the effects in humans. In turn, in each case, there needs to be a debate as to whether there are sufficient data to allow conclusions to be drawn and research to proceed. It would be important to avoid beginning an expensive large-scale, long-term trial without first having the basic information that would ground the rationale for the trial.

OCR for page 123
14 LOOKING TO THE FUTURE PROCESS FOR DRI DEVELOPMENT PANEL DISCUSSION: REFLECTIONS ON WHAT WE HAVE HEARD ABOUT THE PROCESS OF DRI DEVELOPMENT Panel Members: Mary Bush, Jean-Pierre Habicht, Suzanne Harris, Van Hubbard, and Molly Kretsch (later joined by Stanford Miller) The session moderator, Dr. Paul Coates, introduced the panel mem- bers and began the discussion by asking each panelist to offer an opening remark. Panelist Opening Remarks Ms. Bush expressed her belief that the experience of the past 10 years has greatly advanced the DRI development process. The fundamental task will be to ensure that intellectual rigor, as offered by the IOM process, is applied to decision making throughout the next steps for DRI development. She added that timing and resources will be critical considerations. Ms. Bush emphasized the importance of communication, including dialogue at key points between sponsoring agencies and the IOM as well as commu- nication with stakeholders regarding policies and programs based on the DRIs. It was noted that the way in which the scientific concepts are “put together” in the DRI reports is essential to stakeholder understanding—and acceptance—of the process and the reference values and the process used to create them. The process must be clear, transparent, and understood by stakeholders, or the credibility will suffer. Dr. Habicht commented that the scientific work to develop DRIs in a fashion that permits them to be used effectively is still not well advanced. The first task is to deal with the lack of attention to the uses of DRIs. The purpose of DRIs must be clearly enunciated and the objectives of the work related to this purpose. He suggested that consideration of uses earlier in the workshop could have resulted in discussions helpful in remedying the current shortcomings of the DRIs in terms of their applications. His inter- est was in identifying what is necessary and in categorizing applications because approaches for guidance are developed application by application. For example, despite the attention given to the need for information on the distributions of requirements, for many important applications of the DRIs there is no need for information about the requirement distributions. A second example relates to dietary counseling of individuals for whom the dietary guidelines may be the best first-line tool in nutrition counseling; only when this tool reveals special problems does one need to turn to DRIs. In short, the appropriate sequencing of tools has not been systematically ex-

OCR for page 123
14 THE DEVELOPMENT OF DRIs 1994–2004 amined. For this reason a discussion of timing of updates and prioritization of nutrients is premature until systematic, scientific investigations based on considerations of use have been undertaken. He also suggested examining the implications of the imprecision of the coefficient of variation for vari- ous applications as well as the need to reexamine the “false” AIs that exist for some nutrients. Dr. Harris pointed out that there is still much to do to make DRI outcomes and tools as useful as possible for stakeholders. Much can be gained by identifying groups that are interested in such work and willing to take on collaborative efforts, and this would be assisted by a facilitating or organizing mechanism to track the activities. For instance, the efforts being carried out in Europe can be informed by the DRI work, and DRI development can benefit from their input as well. She noted several key topics discussed during the workshop, including separating the DRIs from the dietary guidelines, adding nonessential nutrients important to health to the list of substances considered, and the merits of updating reference values on the basis of single versus groups of similar nutrients. She also noted value in looking for tools to improve the consistency of UL develop- ment. Dr. Harris supported the benefits of risk assessment and its discipline of decision making. Finally, she suggested that changes in the food supply should be taken into consideration and we should ensure the availability of high-quality food composition databases as well as relevant expertise within DRI study committees. Dr. Hubbard stated that the onus is on all of us to be more specific about what we are defining, and then to work to understand how it can be applied best to individuals, groups, special conditions or diseases, and agency/government planning or policy activities. He compared the pro- cess and its principles to the evolution of transportation, which changed in response to both scientific/technical development and needs/purposes. From his perspective, key recognitions that must be embraced include the following: values are best estimates and it is appropriate to assign some probability or level of uncertainty to them as we apply them to individu- als or groups; interactions among nutrients influence their functionality; metabolic changes within the body may occur as we increase the intake of certain nutrients, and this may alter requirements for other nutrients; and the continued focus on chronic disease reduction reduces the specificity of the biomarkers used to set recommendations. As a final point, he suggested that we need to carefully consider the modifications in the process that may be required if we decide to base DRIs on the total population rather than “healthy” persons. Dr. Kretsch addressed the need for research. She highlighted the many research gaps, including endpoints, biomarkers, life stage data, testing the

OCR for page 123
14 LOOKING TO THE FUTURE PROCESS FOR DRI DEVELOPMENT recommendations, and dose–response data near the EAR. She acknowl- edged that although new research is needed, there is no certainty that research will be conducted. She expressed particular concern about the cur- rent opinion that basic research in support of the DRIs is not innovative and will not attract the interest of young scientists. She called for a change in this attitude and stressed the importance of basic nutrition research to reduce uncertainty surrounding the DRIs. She closed her remarks by suggesting it would be useful to review issues related to DRI values and the ability of the population to meet them. Such concerns are very real for practitioners and policy makers alike and warrant attention in the future. General Discussion Dr. Coates opened the discussion to all attendees. However, he first asked panelists if they had additional comments. Guidance for Users One panel member asked whether, as suggested by a presenter in an earlier session, bringing together a group to focus on a particular applica- tion to derive more relevant or specific guidance might be useful. Another responded that we do need to consider the guidance application by ap- plication to improve the guidance offered. A comment was added that the definition of “small groups” is problematic; the two-by-two table specifies individuals and populations, but smaller groups are not addressed. An audience member remarked on the complexity of improving the eating patterns of North Americans given the need to synthesize a great deal of information coupled with the diversity of the food supply. A panel member agreed that it is a challenge to use all of the information to create meaningful dietary patterns, but that it is a needed task so that the guidance given does not require consumers to deal with technical details. Research Priorities A question was asked about approaches for setting priorities given that the study committees had identified research gaps. A panelist suggested it would be useful to identify a coalition of stakeholders who would specify the top five research priorities based on the existing IOM effort to synthe- size the research gaps. An audience member suggested that participants should sponsor junior colleagues to attend future events such as this work- shop in order to foster interest in relevant research.

OCR for page 123
14 THE DEVELOPMENT OF DRIs 1994–2004 Food Composition Databases An audience member expressed concern about some of the shortcom- ings in data that underpin existing tables of food composition. The impact of digestibility and the impact of processing were two examples given. In response, it was noted that for some nutrients the absorbability in the usual North American diet was specified in such tables. One commenter then pointed to the critical need for food databases to keep pace with the food supply so that the proper context for DRI development is available. Funding from For-Profit Groups A question was directed to Dr. Meyers concerning financial support for the DRI project, and whether there was an effort to restrain outside funding from for-profit groups to a level lower than what the Academies allowed. Dr. Meyers indicated that the general policy for the National Academies is that no more than 49 percent of a project’s funds should derive from for-profit entities, to ensure independence, but that these percentages are determined on a case-by-case basis for each project, and the percentage is generally lower. For the DRI project, a small percentage of support was obtained from the Dietary Reference Intakes Private Foundation Fund and the Dietary Reference Intakes Corporate Donors’ Fund. International Collaboration An audience member noted that the European Food Safety Authority (EFSA) was tasked with producing population reference intakes for nutrient substances, including macronutrients. Its report on macronutrients should be available for public consultation next year. She suggested that IOM in- put during this consultation would be greatly appreciated. EFSA also has a second task—to advise the European Commission on how to turn nutrient reference values into food-based dietary guidelines. She contrasted this ap- proach to that used in the United States. A participant then emphasized that international collaboration would provide opportunities to learn from each other and to conserve resources. She asserted that it would be extremely valuable to have at least an informal collaboration to create awareness. Another audience member expressed interest in time lines for next steps given the activities in Europe that would benefit from the DRI work. In response, it was noted that some issues can be resolved more easily than others, but the tasks have to be done correctly and may take time.

OCR for page 123
149 LOOKING TO THE FUTURE PROCESS FOR DRI DEVELOPMENT Derivation of AIs A participant asked Dr. Habicht to clarify his statement on the need to set priorities to reexamine “false” AIs. Dr. Habicht responded that “false” AIs appear to DRI users to have been based on the well-established ap- proach of deriving a recommended intake using a population intake level. He stated that “false” AIs are based on something else. A member of the audience indicated that one approach to developing an AI is to use the me- dian intake of a healthy population, but that AIs had also been developed using an adequate intake level from an experimental study group, and these values are perhaps the ones Dr. Habicht considered to be “false” AIs. Dr. Habicht recommended that the term “AI” be retained for “what it really is” and that another term be used for other approaches. Special Topics of Interest An audience member made a plea to revisit the AI for fluoride, not- ing the disparity between the AIs for children 0 through 6 months and 7 through 12 months. The concern was that the AI for the 0- through 6- month age group was too low to afford the needed public health protection regarding the development of dental caries later in life. Another audience member asked about the practicality of making changes and the possibility of changing values. The response was that if a value cannot be supported, it should be removed or changed. Another participant addressed the question of “other food compo- nents,” such as carotenoids, β-carotene, and flavenoids, or what are re- ferred to as bioactives. He suggested that they currently appear to fall outside the DRI process, but he believed that the task of evaluating the sci- ence and making relevant recommendations for such substances fell within the purview of the IOM. At the end of the discussion, Dr. Coates invited Dr. Sanford Miller to join the panel. He noted Dr. Miller’s considerable experience with DRI development. Dr. Miller offered several comments. He noted that AIs were developed primarily as “placeholders” because no other data were avail- able to allow a recommendation to be made. This approach was taken because failing to issue a reference value fosters the incorrect conclusion that the substance is safe at any level, a conclusion DRI developers wished to preclude. Additionally, he pointed out that the controversy about AIs is broader than it appears. Specifically, he identified the broader issues as follows: What do you do when you lack data? What is the default to be used? Can the data be developed, and how do you present the data given the uncertainty surrounding them? Dr. Miller also focused on the question

OCR for page 123
10 THE DEVELOPMENT OF DRIs 1994–2004 of precision. He indicated that accuracy is needed, but asked how much precision is needed, suggesting an inordinate number of decimal places were used during the development of DRI requirements. Dr. Miller expressed his concern about the possibility of marginal- izing the research that should be at the beginning steps of DRI activities. Although the research needed is often identified as clinical in nature, Dr. Miller countered that the importance of basic biological research cannot be overstated. Such studies outline the physiology involved, and offer a direc- tion for focused human research. He underscored that animal data are not used to extrapolate values for humans, but rather provide information for designing the needed human clinical trials. Finally, Dr. Miller remarked on the considerable number of crosscut- ting issues, the advances that have been made, and the back-and-forth between seeking the advances and returning to basics. He highlighted the importance of “finally beginning to understand the process” and closed by remarking that 14 years ago, this meeting could not have taken place—it is not that the questions have changed, but how we are asking the questions has changed.