Part 2: A Conceptual Foundation for Developing Nutrient Reference Values

Challenges to Harmonizing the Process of Deriving Nutrient Reference Values

An important outcome of the 2017 workshop, Global Harmonization of Methodological Approaches to Nutrient Intake Recommendations: Proceedings of a Workshop (NASEM, 2018a), was the elucidation of specific challenges to achieving a consistent, unified approach to deriving nutrient reference values (NRVs) on a global scale. The challenges identified included the need for:

1. Transparency in the process for deriving NRVs;
2. Standardization among endpoints, particularly related to the risk of chronic disease;
3. Centralized data and information that is accessible to stakeholders globally; and
4. Standardization in the terminology used to identify NRVs.

While this tool kit cannot resolve these challenges, it can provide guidance to countries and regions for utilizing existing tools and resources in a way that facilitates consistency in the approach to deriving NRVs.

Guiding Principles for Setting Reference Values

As described in the Global Harmonization Proceedings of a Workshop and reiterated in the consensus study report, the two core reference values are the average requirement (AR) and the tolerable upper intake level (UL) (see Appendix 2 for definitions). The process for setting the AR, described below should:

1. Be based on the amount of a nutrient estimated to meet the nutrient requirement of 50 percent of healthy individuals in a population;
2. Be established for all essential nutrients and food components that have public health relevance;
3. Include acceptable macronutrient distribution ranges for carbohydrates, protein, and fats that reduce the risk of chronic disease associated with those macronutrients;
4. Consider nutrient–nutrient interactions and quantify them, if possible; and
5. Consider subpopulations with special needs within the defined “apparently healthy” population.

With regard to the first principle, food intake data are only needed to estimate nutrient bioavailability in order to ensure the amount of a nutrient is sufficient to meet the physiologic requirement. Countries or groups that wish to either use or adapt ARs from other sources will need to give further consideration to the usual intake of dietary factors that affect nutrient bioavailability, such as phytate and fiber. The UL is the highest intake of a nutrient that is likely to pose no risk of adverse effects to most individuals in a population. The level of uncertainty in risk of adverse effects must be taken into consideration in setting a UL, bearing in mind that there is variability in uncertainty factors across nutrients.

With the increases in global obesity rates and accompanying chronic diseases, it has become more difficult to define a “healthy population.” More recently, guidelines for the inclusion of chronic disease endpoints in the NRV process have been proposed (NASEM, 2017; Yetley et al., 2017). However, most NRVs for young children and women of reproductive age are aimed at meeting nutrient needs for specific health outcomes, such as growth and cognitive development. Traditionally, chronic disease has not been a priority outcome, although it is a growing concern among low- and middle-income countries.

Current Process for Deriving Nutrient Reference Values

Currently, the process to establish NRVs is initiated by an authoritative body, such as a sponsoring government agency or scientific body. This body, an authoritative national body (e.g., U.S.-Canadian Federal Steering Committee) or a recognized global organization (e.g., the World Health Organization/Food and Agriculture Organization of the United Nations or the European Food Safety Authority) identifies a nutrient for review and then commissions a systematic evidence review. This process comprises four main actions:

1. Accept, Revise Existing, or Conduct New Systematic Review

The starting point for deriving NRVs is to understand the status of the evidence for the nutrient under review. This involves determining whether an existing systematic review is sufficient for the purpose and whether it can be updated. If not, then a new systematic review may need to be commissioned. An expert nutrient review panel is then convened to use the evidence from the systematic review, as well as other relevant evidence that it identifies, to derive the appropriate NRVs. (See the section

“Using the Framework to Derive Nutrient Reference Values” below for detailed information.)

2. Select the Appropriate Methodologic Approach for the Nutrient of Interest

The activities during this phase depend on whether the expert panel chooses to update an existing NRV or establish a new one. A systematic evidence-based review will serve as the basis for deliberation, which includes assessing nutrient requirements for age/gender groups; and establishing findings and conclusions about the nutrient under review.

A de novo literature review will include information about the search criteria, inclusion and exclusion criteria, study quality criteria, data summaries, and correspondence and consistency of the study task with the methodology to be used in the review.

When a relevant systematic review already exists that meets these benchmarks, it should be evaluated to ascertain:

• Whether the questions and criteria identified in the existing review reflect current physiological and biological concerns relative to the nutrient under review, and
• Whether the inclusion/exclusion criteria in the existing review are consistent with and relevant to the Population, Intervention/Exposure, Comparison, Outcome (PI/ECO) elements (described below) for the updated review.

PI/ECO is a technique used to frame and answer a clinical or a health care–related question, for example, in a systematic review. It is also used to develop literature search strategies (Guyatt et al., 2011).

The panel may decide to establish new NRVs for a number of reasons, including substantial differences in the characteristics and/or diet of the population from those on which existing NRVs are based, or the emergence of important new data. New NRVs could be developed for one or more nutrients. Population differences can include various characteristics, which are described below.

If the panel chooses to establish new NRVs, it will need to define its approach for each nutrient, which involves deciding whether to conduct a dose-response assessment or take a factorial approach, which could include balance studies.

3. Assess the Usual Intake Characteristics and Dietary Patterns of the Populations of Interest

Setting relevant NRVs requires an in-depth knowledge and understanding of the dietary components that can affect bioavailability (e.g., phytate or heme iron) in a given population group. National surveys or other large population databases are generally the primary source for population-based intake data. Individual variability in intake can be captured by measuring at least 2 nonconsecutive days of intake on a representative subset of the population. Methods for statistical adjustment of the

distribution of intakes have been described by Carriquiry (1999) and Nusser et al. (1996). Once AR and UL values are determined, the dietary intake data will be used to estimate the prevalence of inadequate (less than the AR) and excessive (above the UL) intakes.

4. Consider Implications or Special Concerns for the Population

The final step involves assessing the local context. This means determining how characteristics of the populations of interest, such as body size, lifestyle, environment, or other factors, may influence the recommended nutrient intake level.

These four actions are embedded in a series of steps illustrated in the following flow diagram (see Figure 1). Additional detail about each step is provided for the process by which expert panels can update or adapt an existing NRV (see Box 1) or establish new NRVs.

The steps in Box 1 outline the tasks that must be completed when updating or adapting existing ARs and ULs to a local context. Differences in physiologic state or body composition that exist between different populations must also be considered. To illustrate, when referring to body weight or energy intake a decision must be made about whether to base the requirement on existing body weights and energy intakes of the population or on reference weights regarded as “healthy” for given height and exercise level.