Nutrient reference values (NRVs) play an essential role in promoting the health and well-being of a population. Conceptually, NRVs are meant to assess the risk of deficiency and inadequacy while avoiding excessive intakes for healthy populations, and to be utilized in planning adequate diets for individuals and populations. The key components of NRVs are the population average requirement (AR) and the safe upper intake level (UL) ¹ . In the 1990s, a group of experts proposed a then novel approach for setting new nutrient intake recommendations. These dietary reference intake (DRI) values were initially adopted in the US, in collaboration with Canada and the UK ¹ .

The World Health Organization (WHO) also provides NRVs (for vitamins and minerals) in a 2004 document, which is based on a 1998 convening joining the WHO and the Food and Agriculture Organization (FAO) of the United Nations ² . However, the DRIs are more widely used. This review will focus on the DRIs developed by the US, UK, and Canada, and not the WHO NRVs, for two main reasons. First, the NRVs set forth by WHO/FAO, despite considering several contextual factors, have not been recently updated for micronutrients. While a few important updates have been made to the DRIs over the years, including calcium and vitamin D, and more recently sodium and potassium ^{1, 3} . In contrast, the WHO NRVs have had updates only in sodium and potassium, as well as in calcium supplementation ^{4– 6} . Second, the DRIs and NRVs are similar for most micronutrients in terms of AR (but not UL). A few micronutrients have DRIs but not NRVs (in general or in pregnancy), such as choline, vitamin E, phosphorus ^{2, 7– 9} .

Nutrient reference values are especially important during pregnancy. There is a biological imperative for increased nutrient intake, and there are well-documented maternal and fetal consequences of inadequate nutrition during the periconceptional, pregnancy, and postpartum periods ¹⁰ . Further, NRVs in pregnancy inform the composition of “prenatal vitamins”, or multiple micronutrient supplements (MMS), used in pregnancy to promote positive pregnancy outcomes ¹¹ . The commonly used UNIMMAP supplement for pregnancy includes 15 vitamins and minerals at the level recommended by the DRIs.

Yet, there is growing recognition of the limited extent to which the DRI study populations represent all sub-populations, including pregnant and lactating women ¹ . Women have historically been underrepresented in medical research. For example, drug trials to date have been conducted almost exclusively in young men ¹² , and a recent review of clinical trials (including nutraceutical) in India found that less than 2% included pregnant women ¹³ . This is problematic because there is likely to be sexual dimorphism in the physiology, metabolism, and related toxicity and efficacy of supplements and drugs. Further, there are clear ethical imperatives to include pregnant women in research because women deserve access to well-studied interventions ¹⁴ . We hypothesize that nutrition science and the related evidence-based guidelines have similarly suffered from excluding women and pregnant women from research.

Meta-research, or meta-science, is an evolving field focused on systematic evaluation of the way science is produced and used ¹⁵ . There are five broad areas of focus for meta-science including: the way research is performing, or methods; the way people communicate about research, or reporting; the way we verify research, or reproducibility; the way we conduct peer review, or evaluation; and the way we reward research, or incentives ¹⁵ . Meta-research is well suited to measure whether specific subpopulations are excluded from biomedical research and whether novel research methods are equally applied to problems or conditions that affect women, underrepresented minority populations, and people living in the global south.

We aim to formally review the methods of research studies used to inform the DRIs and identify opportunities to improve how we fund, perform, evaluate, and incorporate nutrition science for improving public health. The objective of this meta-science study is to summarize the population and nutrition-science methods used in studies informing the nutrient reference values for women, and for pregnant and lactating women.

This review is focused primarily on assessing the population and methods used in the studies informing the development of DRIs:

Secondary research questions include:

This review study will not assess the quality of the research or risk of bias in each study. Nor will it assess whether the interpretation and conclusions of the studies are accurate or appropriate.

The search process will begin with identifying the indicators considered by the committee to establish the DRI. All references in the “Selection of Indicators for Estimating the Requirement for [nutrient]” section will be abstracted. We will also abstract references in the “Findings by Life Stage and Gender Group” Section, specifically for the “Adults”, “Pregnancy”, and “Lactation” sub-sections. Finally, we will abstract references in the “Tolerable Upper Intake Levels” section. The full reference (as listed in the DRI document) and the section in which it appeared will be recorded.

There are two exceptions in the search process requiring slight adjustments to the search strategy. First, for vitamin B2 (riboflavin), an additional section will be included, “Approaches for Deriving the Estimated Average Requirement” because this section provided an overview of B2-related references used to determine the EARs. Second, the latest (2011) calcium and vitamin D report has a different layout of the chapters and sections from other reports. We will abstract references from three chapters in this report: Chapter 4, “Review of Potential Indicators of Adequacy and Selection of Indicators: Calcium and Vitamin D” (Tables 4–8, 4–9, 4–11, 4–12 and the preeclampsia section will be our focus); Chapter 5, “Dietary Reference Intakes for Adequacy: Calcium and Vitamin D”; and Chapter 6, “Tolerable Upper Intake Levels: Calcium and Vitamin D”. To ensure the completeness of the data, we will also abstract references in the “life stages” section of the 1997 calcium and vitamin D report, which determined adequate intakes (AIs) but not estimated average requirements (EARs) for these two nutrients (EARs were set in the 2011 report).

The following five documents will be reviewed during the search:

Three documents will not be reviewed, given that the focus of this meta-research is micronutrient DRIs:

Two researchers will be assigned to each nutrient list and will independently screen the references: one researcher as the main screener and the other for quality control. The rapid screening phase will involve reviewing the DRI document and the study title to determine whether the study meets the inclusion criteria for the study. The product of this screening step will be a list of final records for full-text review and data extraction.

Inclusion criteria

Exclusion criteria

We will abstract data for all studies that pass through the screening phase. Data will be entered in a study-specific data extraction form, and the following data will be collected.

Administrative information, study identification

Study methods

Human population

Non-human subjects

After the completion of full-text data extraction, qualitative synthesis will be conducted for all included studies, separately for each micronutrient (and combined for all selected nutrients as needed). The synthesis will prioritize the research questions in terms of population characteristics and methods used in the included studies. We will also import the cleaned dataset into a statistical software (R version 4.0.4, R Core Team, Vienna, Austria) for analysis ¹⁶ . We will conduct univariate descriptive analysis (e.g., summarizing number of included studies for each nutrient); between-group comparisons (e.g., to compare findings between two nutrients, or between two sub-groups), including the Student’s t-test (or non-parametric test such as the Mann-Whitney U test), chi-squared test; as well as trend test (e.g., to examine annual trends of the publications for individual and all included nutrients).

We will publish the paper in an open access scientific journal. Depending on the amount of evidence we eventually review and generate, there may be more than one manuscript as a result of this project. We will upload our primary (raw) data upon completion of the review to an approved online repository (e.g., Figshare or Harvard Dataverse). We will then share the repository DOI with the manuscript.

Figure 2 shows the proposed timeline of the study. At time of publication, we have now completed the compiling of references (pulled citations from the DRI reports), and are currently screening the references. In addition, we have pilot-tested the entire review process for one micronutrient (vitamin B12).

No underlying data are associated with this article.