8

The Process for Establishing Chronic Disease Dietary Reference Intakes

Chapters 1, 2, and 3 briefly explain the history of the Dietary Reference Intakes (DRIs) in the United States and Canada, the process to establish them, and the challenges that nutrition researchers face when assessing the associations between nutrients and chronic diseases. Chapters 4, 5, 6, and 7 offer recommendations and guiding principles to address the methodological and conceptual challenges when recommending chronic disease DRIs.

Chronic disease, however, is only one of several types of indicators that are reviewed through the DRI process. Other key indicators are those for adequacy and toxicity, which are not the focus of this report but will also continue to be reviewed and assessed so that Estimated Average Requirements (EARs), Recommended Dietary Allowances (RDAs), and Tolerable Upper Intake Levels (ULs) can be established. The establishment of DRIs occurs with the collaboration of a number of stakeholders, and within that collaboration, each stakeholder has a specific role to contribute. For example, the Canadian and U.S. Steering Committees identify the nutrients and questions to be addressed. The systematic review team conducts the systematic review in consultation with a technical expert panel. The DRI committee—which ideally will include some members of the technical expert panel—receives and assesses the systematic review, makes conclusions based on the certainty in the evidence, and makes DRI recommendations based on one or more indicators (i.e., adequacy, toxicity, and chronic disease risk). The goal of this chapter is to address two questions related to specific aspects of the DRI process and for each, recommend one of the options offered in Options for Basing Dietary Reference Intakes (DRIs) on Chronic Disease Endpoints: Report from a Joint US-/Canadian-Sponsored

Working Group (i.e., the Options Report) (Yetley et al., 2017). The first question deals with how to incorporate the deliberations and recommendations regarding chronic disease DRIs into the already existing DRI process, which relies on the work of a National Academies of Science, Engineering, and Medicine (the National Academies) committee (i.e., DRI committee). The second question addresses the nature of the initial question that will drive the systematic review and ensuing tasks. This committee provides a justification for selecting one of the options to answer these two questions but notes that tasks such as deciding when and how to select a nutrient/chronic disease for review are out of the scope for this report, as mentioned in Chapter 1.

CONVENING A DRI COMMITTEE AND THE ROLE OF DRI COMMITTEES

As Chapter 2 explains, since 1938 and 1941, nutrient reference values have been issued in Canada and the United States, respectively. In the United States, the nutrient reference values (both the original RDAs and the more recent DRIs) were published by the Food and Nutrition Board (FNB) of the National Academy of Sciences. In the current DRI process—initiated in the early 1990s and then modified for the latest DRIs on calcium and vitamin D in 2011—the National Academies convenes an expert committee to develop DRIs for a group of related nutrients (see Table 2-1, Chapter 2). In addition to following the National Academies’ policies and procedures, committees must comply with provisions of section 15 of the Federal Advisory Committee Act¹ so that U.S. government agencies can use the recommendations provided.

When they are appointed, DRI committees are given a statement of task, which is a description of the objective of the project and the specific charge to the committee. The statement of task is typically developed by the governments of the United States and Canada, which also have been the sole funders of the work of DRI committees, and in consultation with the FNB. DRI committees are composed of scientists in the relevant disciplines that are needed to complete the specific task (e.g., human nutrition, epidemiology, toxicology) and are convened based on a selection process that considers suggestions by stakeholders and potential conflicts of interests and biases following the policies of the National Academies. DRI committees are supported by FNB staff, who coordinate their work and ensure completion of the task and the publication of a report containing the DRIs as well as the scientific rationales, in the style of the National Academies reports. It generally takes 1 to 2 years for a DRI committee to complete

___________________

¹ Federal Advisory Committee Act, 5 U.S.C. § 15.

its work. As mentioned in Chapter 1, the committee developed its recommendations and guiding principles in the context of the process shown in Figure 1-2 in which formal systematic reviews are conducted by an outside contractor before the formation of the DRI committee. Current practice is that, for each set of related nutrients, one committee addresses all indicators (i.e., adequacy, toxicity, and, possibly, chronic disease outcomes) to establish DRIs. As Chapter 2 indicates, for a single nutrient one or more DRIs may be established, and many nutrients have an EAR and an RDA (or an AI) and a UL. If adequate data are available, the DRIs may incorporate chronic disease considerations (IOM, 2003). Chapter 2 lists all the DRIs established to date and the reports by the National Academies in which they have been published.

As Chapters 6 and 7 explain, future DRI committees that will consider chronic disease DRIs not only will have to assess whether a DRI can be recommended for a particular chronic disease, but also will have to balance harms and risks related to all relevant indicators. Although risks and harms also were considered in the past, the scenarios will likely become more complex as committees formally consider chronic disease indicators and more evidence becomes available. For example, future DRI committees may have to consider situations where one nutrient or food substance (NOFS) may be associated with more than one chronic disease but within different ranges of intake. Other scenarios can be anticipated in which a range of intakes for one nutrient may provide benefits for a chronic disease outcome but may overlap with the UL.

CONVENING OF DRI COMMITTEES: OPTIONS AND JUSTIFICATION

Box 8-1 shows the options presented in the Options Report with regard to the DRI process. Option 1 was to continue to use a single DRI development process and option 2 was to create two separate processes for developing DRIs. The committee recommends a variation of option 1, such that for each set of NOFSs under review, a single DRI “parent” committee would be formed but it could have two subcommittees focused on different questions. If two subcommittees are formed, the parent committee would then integrate the recommendations of each subcommittee into a single report.

The committee chose not to use option 2 because the Federal Advisory Committee Act rules, which apply to committees of the National Academies and are meant to minimize external influences, require that committee members and staff keep deliberations confidential. Two separate DRI processes would not allow for sufficient exchange of ideas between the two committees. Because of the need for coordination of the recommendations,

and in particular the need to consider harms and benefits with regard to all health indicators, one DRI committee would be a better choice. This committee structure will allow for full exchange of ideas and deliberations among all members.

Because of the importance of drawing on historical knowledge and experience with the process, the National Academies would be the logical organization to coordinate the process. When enough scientific evidence for an update of the DRIs exists for a given set of NOFSs, and as requested

by sponsors,² the FNB would convene an ad hoc consensus committee (parent committee) that would be responsible for establishing the DRIs. A wide range of expertise would need to be represented on the committee to accommodate the interdisciplinary approach needed to set standards for reducing the risk of chronic disease as well as the narrower perspective of preventing nutrient deficiencies and toxicities. As Guiding Principle 9 states (see Chapter 6), “Particular elements of needed expertise will be guided by the general scientific question(s) and specific questions and will generally include nutrition science, scientific study design and analysis, public health, biostatistics, nutrition and chronic disease epidemiology, disease pathogenesis, and evidence review conduct.”

It should be noted that not all nutrients have been associated with a chronic disease. Thus, some DRI committees would not have a chronic disease DRI subcommittee. When two subcommittees are formed, the FNB staff working with the parent committee would coordinate the work of the two subcommittees so that a full exchange and integration of ideas between the subcommittees could occur. Various strategies would be implemented to ensure this cross-fertilization of ideas, such as designating some members to participate in both subcommittees and the parent committee, organizing meetings that all members would attend, and sharing evidence tables (see Chapter 6) and other documents.

When two DRI subcommittees are formed, they would approach the development of DRIs for a nutrient or a group of nutrients as separate processes, though they would use the same formally conducted systematic evidence review to complete their work.³ One subcommittee would recommend DRIs for all 22 life-stage groups that focus on the requirements for specific nutrients to ensure nutrient adequacy and prevent nutrient deficiency symptoms as well as prevent toxicities. The second subcommittee would examine the certainty of the scientific evidence on the relationship between the nutrients under review and chronic disease outcomes, keeping in mind the distinction that chronic disease DRIs are desirable but not essential. The evidence review and, ultimately, the certainty in the data regarding the intake-response relationship (see Chapter 7) would dictate whether it is necessary to recommend DRIs for one or more NOFSs, including NOFSs individually or in combination that may reduce the risk of a chronic disease of interest. If insufficient certainty existed in the evidence

___________________

² As described in Chapter 6, once the general nutrient and health outcomes of interest have been identified, an important next preliminary step is to conduct a scoping review, that is, to sample the scientific literature and determine whether sufficient studies exist to conduct a full systematic review.

³ Although it is assumed that both subcommittees will have access to the same systematic review(s), this committee is not commenting on the process of setting nutrient reference values for adequacy or toxicity, per the statement of task.

to develop DRIs for reducing risk of chronic disease for the nutrients under review, the second subcommittee’s process would stop. If sufficient evidence did exist to recommend chronic disease DRIs for any of the life-stage groups, all members would deliberate their conclusions about DRIs for all indicators (i.e., adequacy, toxicity, and chronic disease risk), consider benefits and harms, and recommend the chronic disease DRIs, incorporating all needed documentation and explanations in the report (see Guiding Principles). If recommended for any of the life-stage groups, these chronic disease DRIs could be defined as proposed in Chapter 7 (see Table 7-1), that is, as acceptable range of intake, range of beneficial increased intakes, or range of beneficial decreased intakes.

An alternative to the committee structure above would be to convene only one parent committee, with all the necessary expertise, but no subcommittees, to establish all DRIs. However, this committee concludes that the process to derive DRIs based on adequacy and toxicity would not be adequate to derive chronic disease DRIs. Although the general approach to establishing DRIs would be consistent with the same general risk framework (see Annex to Chapter 1), developing chronic disease DRIs presents unique challenges (see Chapter 3) that can only be adequately addressed by applying specific methodological and statistical models (see Chapter 7). In fact, although the same systematic review(s) might be used to compile the relevant information for all indicators, the type of evidence used as a basis for chronic disease DRIs will differ from the type of evidence used to establish DRIs for adequacy and toxicity because of the unique features of the relationships between nutrients and chronic diseases. A few examples are:

• When dietary factors influence the risk of chronic disease, DRIs will likely be expressed in ranges rather than a single number. In some cases, a DRI recommendation regarding chronic disease endpoints would most appropriately involve a ratio of nutrients.
• Multiple factors (including foods and nutrients) are associated with the risk of chronic disease. The current EAR, RDA, and UL models that focus on biological outcomes or functions for a single nutrient will not work well for chronic disease endpoints that have multiple risk factors.

STARTING-POINT ISSUES: OPTIONS AND JUSTIFICATION

As the Options Report indicates, the current starting point for establishing DRIs is individual or small groups of NOFSs (Option 1 in Box 8-2). However, given the number of NOFSs that might affect a chronic disease, the Options Report presented a second potential approach, whereby DRI

committees would assess the evidence on all NOFSs related to a particular chronic disease.

The committee deliberated about the potential advantages of this second option as an alternative to the current process, including the fact that this option could be, in principle, more helpful for those who have higher risk of developing a particular chronic disease. However, the committee found that continuing with the current approach of recommending DRIs for individual or small groups of related NOFSs has advantages, at least for now. First, relatively little experience has been accrued in developing DRIs based on chronic disease, so it may be premature to change the current process before additional experience is gained. Second, the current scientific literature and study designs have tended to explore relationships substance by substance; that is, individual studies are more likely to document how a particular NOFS is related to various diseases rather than studying all the NOFSs that are related to a particular chronic disease. Although the evidence base on how dietary patterns relate to chronic diseases is growing, it is difficult to use this evidence to develop DRIs for each individual NOFS.

In addition, the 2015 Dietary Guidelines Advisory Committee⁴ already addressed the relationship between dietary patterns and chronic disease risk. As the research findings increase, it may be possible to identify key nutrients within dietary patterns that are influencing chronic disease risk. When that occurs, this approach to review the evidence (or starting point) for recommending chronic disease DRIs could be reconsidered.

Finally, and perhaps most importantly, the fact that some NOFSs might contribute to more than one chronic disease implies that, if this second option were selected, balancing of harms and benefits would be more challenging, if not impossible. The committee would lack essential information if, for example, an NOFS range reduces the risk of the chronic disease in question but affects the risk of another disease in the opposite direction.

For the reasons noted above, the committee recommends option 1, that is, when sufficient evidence exists to develop chronic disease DRIs, DRIs should be recommended for one or more NOFSs that are interrelated in their causal relationships with chronic disease(s). As knowledge about relationships between NOFSs and chronic diseases advances, the Canadian and U.S. DRI Steering Committees might expand the specific questions to consider not only the effects of the NOFSs of interest themselves, but also NOFSs that act as effect modifiers. For example, questions about the potential effects of calcium, magnesium, and potassium in the association between sodium and blood pressure could be included in the systematic review protocol for DRIs (see also Chapter 7, “Were Interactions with Other Nutrients Considered?).

Since 1994, DRIs have served as the foundation for nutrition policies and guidance for individuals and groups in the United States and Canada. The Dietary Guidelines for Americans and the Canadian Food Guide, for example, draw on information from the DRI reports. DRIs have many other important uses, such as providing benchmarks for monitoring dietary intake of populations, evaluating the quality of government food assistance programs, planning foods and diets for military personnel, and planning and monitoring other nationwide health programs. However, in the past DRIs were based on reaching adequacy and minimizing the potential for toxicity. As populations have changed their diet habits and more information about the prevalence of chronic disease and its risk factors has become available, there is a need to explore how nutrients contribute to chronic

___________________

⁴ The Dietary Guidelines Advisory Committees review the body of scientific and medical evidence in nutrition and prepare an Advisory Report for the Secretaries of the U.S. Department of Health and Human Services (HHS) and the U.S. Department of Agriculture (USDA) every 5 years. The Advisory Report provides an evidence base for HHS and USDA as the departments update the Dietary Guidelines for Americans (https://health.gov/dietaryguidelines/purpose.asp; accessed May 10, 2017).

disease and whether their specific levels can be determined to ameliorate the risk of chronic disease. Decisions about chronic disease DRIs, however, are fundamentally different from decisions concerning adequacy and toxicity DRIs. Differences lie in the nature of the health outcomes (e.g., the long-term nature of chronic disease), the scientific data available, and the methodologies required to analyze the data. Differences also exist in the expertise needed. For these reasons, it is necessary to develop some guiding principles and recommendations that take these differences into account. Although the United States and Canada have many years of experience in setting DRIs, integrating chronic disease as a focus is a fairly recent task; therefore the recommendations in this report should be revisited in the future as more practice and knowledge is gained. As chronic disease DRIs are set, there will be a need to develop guidance, possibly separate from the DRI reports themselves, on how these new DRIs could be used in dietary assessment and planning, especially in complex situations, such as a single NOFS with DRI ranges to lower risk of different chronic diseases.

REFERENCES

This page intentionally left blank.