Dietary Reference Intakes: resuscitate or let die?

The Dietary Reference Intakes (DRIs) are the most widely used values globally for recommended intakes of essential nutrients. Their US origins are the Recommended Dietary Allowances (RDAs) developed in 1941 by the newly established Food and Nutrition Board. The RDAs were revised 9 times until the final 10th edition was published in 1989. Corresponding Canadian nutrient reference intakeswere first created in 1939 by the CanadianCouncil on Nutrition. They were revised multiple times, renamed RecommendedNutrient Intakes in 1983, and last revised byHealth Canada in 1990. In themid-1990s, the Food and Nutrition Board and Health Canada agreed to collaborate on the development of a more scientifically rigorous, comprehensive approach to recommendations for essential nutrient intakes, resulting in a new paradigm called the Dietary Reference Intakes. The first DRIs were issued in 1997 for calcium, phosphorous, magnesium, vitamin D, and fluoride (1). During the following 8 y, the DRIs for the remaining essential nutrients were released in 5 additional reports. The last DRIs, for water, potassium, sodium, chloride, and sulfate, were issued in 2005 (2). Although additional companion reports have been issued since then for dealing with guiding principles, applications, risk assessment, and proposed additional definitions, the DRIs for essential nutrient intakes were not reassessed and revised until the 2011 release of updated DRIs for vitamin D and calcium (3). Thus, except for vitamin D and calcium, none of the remaining nutrient DRIs have been updated in more than a decade. Furthermore, only the recent revisions of the vitamin D and calcium DRIs were developed by using the tools of formal systematic evidence review and analysis. Earlier DRIs for the remaining nutrients were reached primarily by consensus evaluation of available data at the time by unpaid volunteer committees. Moreover, in the past decade, the nutrition community has voiced considerable interest in the potential development of DRIs for food constituents not traditionally classified as essential nutrients and in further assessing a more-formalized, systematic approach for including chronic disease endpoints in formulating DRIs for individual nutrients. The Food andNutritionBoard has recognized the need for revision of the remaining DRIs and has struggled for some years to find a justifiable and practical approach to do so. A periodic update is essential to maintain the credibility of the DRIs. Notably, by an act of Congress, the Dietary Guidelines for Americans are updated every 5 y, but no regular update of the DRIs, which are meant to inform the guidelines, occurs. Furthermore, the knowledge that evidence will be reviewed regularly provides an incentive for investigators to conduct research that could contribute to improved DRIs. As former members of the Food and Nutrition Board, both of the authors participated in some of these discussions. In this issue of the Journal, Brannon et al. (4) propose a welcome framework based on “evidence scanning” to facilitate the identification of nutrients for which significant new data could justify reassessment of their DRIs. This would also identify nutrients for which little or no new data can be found that might inform a reassessment. The framework is constructed around a risk assessment approach described and tested in the epidemiologic literature (5, 6) to identify a “signal” for updating systematic reviews. This approach aims to balance the ideal goal of regularly conducting formal systematic reviews for each essential nutrient with the reality of limited funds and personnel available for conducting such reviews. Brannon et al. (4) tested their thesis with the use of literature published in the past decade on thiamin and phosphorus. In neither case was there sufficient new evidence to justify a formal systematic review to update the previously published DRIs for these nutrients. The basic issues here are relatively straightforward, but surely not simple. Few would argue that the existing DRIs need either updating or confirmation that the published values remain appropriate. Few would disagree that the consensus approach taken to establish the existing DRIs no longer satisfies the highest standards for evidence-based decisions. Fewwould disagree thatmany of the tools for systematic reviews and formally grading evidence quality, risk of biases, study heterogeneity, and the like were not applied to earlier DRI recommendations. Consequently, few would likely object to formal systematic assessment of the evidence available to establish specific nutrient DRIs conducted with the input of those with deep knowledge of the topic. Nor should there be much disagreement with the position that evaluation of the evidence requires evaluation of all of the evidence, not just recent evidence. The proper way to formally and comprehensively review the evidence is to interrogate the entire literature, including the data used to establish the earlier DRIs. In other words, in the case of the DRIs for nearly all nutrients,