Annette Dickinson, Ph.D.05.01.04
For the past decade, the Institute of Medicine's (IOM) Food and Nutrition Board (FNB) has been evaluating current knowledge of nutrition, and now the time has come to consider putting the new knowledge into practice. Numerous expert committees on particular nutrients have met, debated and studied the evidence and have published six large books containing current Dietary Reference Intakes (DRIs) for those nutrients, beginning in 1997 and concluding just this year. This effort represents a new and expanded approach to establishing Recommended Dietary Allowances (RDAs).
In 1943, RDAs were established and revised periodically as scientists gained new understanding about the levels of nutrients required to avoid deficiency and promote good health. The tenth and last edition of the older, simpler RDA series was published in 1989. The current volumes include not only RDAs but also Estimated Average Requirements (EARs), Adequate Intakes (AIs) and Tolerable Upper Levels of Intake (ULs) for vitamins and minerals, as well as electrolytes and the macronutrients (fat, protein, carbohydrates).
While the DRI reports represent a new level of sophistication in evaluating nutritional needs, will they make a practical difference to the food industry, marketers of dietary supplements or consumers? Believe it or not, they will, although whether that impact will be positive or negative for the consumer remains to be seen.
Since 1941 in the U.S., at least some foods have been labeled with nutrition information based on standards that were current at the time. The first vitamin and mineral standards were called Minimum Dietary Requirements (MDRs), and labels of fortified foods and dietary supplements were required to show what levels of vitamins or minerals had been added to the products, as a percent of the MDR.
A few decades later, FDA developed new regulations on nutrition labeling and adopted a new reference standard called the U.S. Recommended Daily Allowance (U.S. RDA). This was based on the 1968 version of the RDAs, but simplified for easy use in labeling, with a single nutrient value chosen as the label reference for each nutrient. This single value was meant to cover the needs of all men and women, teenagers and children ages four or older. Since the scientific version of the RDAs established separate values for males and females in a large number of age groups, the question was which number to pick as the reference value for nutrition labeling. The decision was made to pick the highest number in order to provide the public with a target for nutrient intake that would cover the needs of all. If the RDA for vitamin C ranged from 40 mg to 60 mg for various age and gender groups, then 60 mg would be the label reference value, and a product containing 30 mg of vitamin C would have a label showing that it provided 50% of the U.S. RDA for vitamin C, per serving.
In 1990, FDA was considering a major revision of nutrition labeling, and the agency floated the notion of reducing the label reference values for vitamins and minerals. Instead of a population-coverage approach based on the highest RDA for each nutrient, FDA suggested calculating a population-weighted mean. To take a fictitious but simple example: If children with an RDA of 40 mg for vitamin C made up half the population and adults with an RDA of 60 mg made up half the population, then the label reference value for vitamin C would be 50 mg-less than the RDA for adults. This whole idea was strongly opposed by some academics and industry representatives who thought it was bad public policy to establish an official value for nutrition labeling that underestimated the desirable target intake for a substantial proportion of the population. FDA eventually abandoned this approach, and the final nutrition labeling regulations issued in 1993 incorporated the higher label reference values for vitamins and minerals
Since the DRI effort began, FDA has been saying it would consider whether to revise nutrition labeling based on the new scientific information that would emerge in the DRI reports. FDA also contracted with the Institute of Medicine (IOM) to convene a special committee to provide advice on how best to incorporate the new DRIs into nutrition labeling. The labeling committee was chaired by the highly respected Irwin Rosenberg of Tufts University. In December last year, the DRI labeling committee published its report after two years of deliberations. Unfortunately, it harks back to FDA's old idea of lowering the label reference values. Even more shocking is the fact that the committee recommends lowering the values below what FDA had proposed in 1990.
Now that the FNB has spent a decade establishing new RDAs for many nutrients, carefully evaluating the data and increasing the recommendations for some nutrients for some age groups, the labeling committee recommends tossing all that aside and basing nutrition policy and labeling instead on the Estimated Average Requirement (EAR). This is the amount of a nutrient that is supposed to meet the actual nutritional need for half of a specific population. By definition, this amount is below the actual nutritional needs of half that population. Once an EAR has been established, a safety factor is added to it, increasing it by two standard deviations (about 20% for most nutrients). This higher number is the RDA, which is calculated to cover the needs of 97.5% of the population. Why should nutrition labeling and nutrition policy in the U.S. be targeted toward meeting the needs of only half the population, instead of meeting the needs of almost all?
But waitit gets worse. Not only does the committee recommend using the EAR instead of the RDA-it recommends using a population-weighted EAR. So what?
Well, let's look at vitamin C as an example. Vitamin C was one of the nutrients whose RDA was increased in the new DRI reports. The RDA for adults had been 60 mg, but the new reports increased it to 75 mg for women and 90 mg for men-a 50% increase for men. If nutrition labeling were to be based on the highest RDA, the new reference value for vitamin C would be 90 mg. But instead, the IOM's committee suggests using the EAR, which is only 60 mg for women and 75 mg for men. The values are even lower for children and teens, so the population-weighted reference value would remain stuck at or near 60 mg. Thus, under the labeling committee's recommendations, absolutely no public benefit is derived from the FNB's decision to revise the vitamin C recommendations upward. (See Table 1)
Iron provides another example. The RDA for iron is 18 mg for women of childbearing age, 8 mg for adult men and postmenopausal women and somewhere in between for children and teens. The EAR is lower-8 mg for women of childbearing age and between 4 mg and 8 mg for everybody else. The current label reference value for iron is 18 mg, and the new one under the scheme proposed by the IOM's labeling committee would be a mere 6.1 mg-only one-third of the RDA for women of childbearing age and in fact below the EAR for those women. (See Table 1)
The FNB recommended higher intakes of vitamin D for the elderly, but these increases would be lost in the population-averaging approach. Supplemental vitamin B12 was also recommended for the elderly-in an amount that is greater than the new reference value, so a breakfast cereal or a dietary supplement labeled to contain 100% of the DV for vitamin B12 would not, in fact, contain the supplemental amount now recommended for the elderly. This would be a shame, brought about by a wrongheaded approach to nutrition labeling, if FDA were to adopt the labeling committee's recommendations.
The DRI reports all include a basic discussion of the purposes of the RDA as compared to the EAR. The RDA is clearly identified as the target intake for the individual. Thus, for purposes of nutrition education, it is the RDA that should be the focus of communications to the consumer. Nutrition labeling is clearly a major form of communication to the consumer regarding target levels of nutrient intake. The percent DV shown on the label is not an abstract figure. It gives the consumer information about the fraction of the daily need for each nutrient provided by the product. For some products, such as fully fortified breakfast cereals and dietary supplements formulated to provide 100% DV for a variety of nutrients, the label value is not only a tool for consumer education but is also the basis for product formulation.
If the sole purpose of nutrition labeling were to compare one food to another, either the RDA or the EAR would do, and the committee recognizes this. Why then opt for a marginal standard rather than a more generous one? Could it be so that foods "look better" to consumers? Surely this was not the motive, but it would definitely be the outcome. Foods that contain at least 10% of the reference value for a particular nutrient can be called "good sources" of that nutrient, and foods that contain 20% can be called "excellent sources." If the reference standard is lowered, as the committee recommends, then more foods will meet these requirements.
The labeling committee expresses concern about the potential for over-fortification of the food supply, and it appears that one objective is to decrease overall levels of fortification. For example, a breakfast cereal that now provides 100% of the DV for a variety of vitamins and minerals would, in the future, contain less of each nutrient and still be able to claim 100% of the DV. The same would apply to vitamin and mineral supplements. Would people be pleased to know that a bunch of scientists are proposing to reduce the amounts of nutrients in their fortified foods and multivitamins, or would people be outraged at this subterfuge? The latter seems more likely.
This might all be a tempest in a teapot if the RDA had no real significance to the consumer and if replacing it with the EAR were purely a policy decision with little or no real impact on consumer health. However, the DRI reports repeatedly emphasize that an individual should strive to consume RDA levels of vitamins and minerals. The following is a paraphrase of the advice for the individual from the DRI reports (emphasis in original):
If a person's usual dietary intake is at least as high as the RDA, there is little likelihood that the intake will be inadequate.
If a person's usual intake is in between the RDA and the EAR, there is a great deal of uncertainty about whether that intake is inadequate-it will be inadequate for 3-50% of the individuals in that person's age and gender group. If intakes between the RDA and the EAR are maintained over a prolonged time period, a person might demonstrate signs of inadequacy.
If a person's usual intake is less than the EAR, there is a high likelihood that intake is inadequate.
Since the DRI reports recognize that intakes near the EAR may lead to signs of inadequacy over time, why would any policymaking agency select the EAR as the basis for nutrition labeling and related efforts in nutrition education? The rational choice is to select the RDA as the appropriate standard and to select the highest RDA rather than averaging out the needs of vulnerable groups.
The FDA would be well advised to set the labeling committee's recommendations aside and stick with the tried and true method of choosing a population-coverage approach to nutrition labeling. The highest RDAs have served consumers well as targets for healthy nutrient intakes for 30 years. FDA should continue to use high values as targets for nutrition policy and nutrition labeling for at least 30 more years. This means using RDAs rather than EARs as the basis for the reference value and using the highest value for each nutrient rather than applying a population-weighted average. NW
About the author:
Annette Dickinson, Ph.D., is president, Council for Responsible Nutrition (CRN), Washington, D.C. She can be reached at 202-776-7929; Fax 202-204-7980; E-mail: annette@crnusa.org.
In 1943, RDAs were established and revised periodically as scientists gained new understanding about the levels of nutrients required to avoid deficiency and promote good health. The tenth and last edition of the older, simpler RDA series was published in 1989. The current volumes include not only RDAs but also Estimated Average Requirements (EARs), Adequate Intakes (AIs) and Tolerable Upper Levels of Intake (ULs) for vitamins and minerals, as well as electrolytes and the macronutrients (fat, protein, carbohydrates).
While the DRI reports represent a new level of sophistication in evaluating nutritional needs, will they make a practical difference to the food industry, marketers of dietary supplements or consumers? Believe it or not, they will, although whether that impact will be positive or negative for the consumer remains to be seen.
History
Since 1941 in the U.S., at least some foods have been labeled with nutrition information based on standards that were current at the time. The first vitamin and mineral standards were called Minimum Dietary Requirements (MDRs), and labels of fortified foods and dietary supplements were required to show what levels of vitamins or minerals had been added to the products, as a percent of the MDR.
A few decades later, FDA developed new regulations on nutrition labeling and adopted a new reference standard called the U.S. Recommended Daily Allowance (U.S. RDA). This was based on the 1968 version of the RDAs, but simplified for easy use in labeling, with a single nutrient value chosen as the label reference for each nutrient. This single value was meant to cover the needs of all men and women, teenagers and children ages four or older. Since the scientific version of the RDAs established separate values for males and females in a large number of age groups, the question was which number to pick as the reference value for nutrition labeling. The decision was made to pick the highest number in order to provide the public with a target for nutrient intake that would cover the needs of all. If the RDA for vitamin C ranged from 40 mg to 60 mg for various age and gender groups, then 60 mg would be the label reference value, and a product containing 30 mg of vitamin C would have a label showing that it provided 50% of the U.S. RDA for vitamin C, per serving.
In 1990, FDA was considering a major revision of nutrition labeling, and the agency floated the notion of reducing the label reference values for vitamins and minerals. Instead of a population-coverage approach based on the highest RDA for each nutrient, FDA suggested calculating a population-weighted mean. To take a fictitious but simple example: If children with an RDA of 40 mg for vitamin C made up half the population and adults with an RDA of 60 mg made up half the population, then the label reference value for vitamin C would be 50 mg-less than the RDA for adults. This whole idea was strongly opposed by some academics and industry representatives who thought it was bad public policy to establish an official value for nutrition labeling that underestimated the desirable target intake for a substantial proportion of the population. FDA eventually abandoned this approach, and the final nutrition labeling regulations issued in 1993 incorporated the higher label reference values for vitamins and minerals
Since the DRI effort began, FDA has been saying it would consider whether to revise nutrition labeling based on the new scientific information that would emerge in the DRI reports. FDA also contracted with the Institute of Medicine (IOM) to convene a special committee to provide advice on how best to incorporate the new DRIs into nutrition labeling. The labeling committee was chaired by the highly respected Irwin Rosenberg of Tufts University. In December last year, the DRI labeling committee published its report after two years of deliberations. Unfortunately, it harks back to FDA's old idea of lowering the label reference values. Even more shocking is the fact that the committee recommends lowering the values below what FDA had proposed in 1990.
The Impact of Using an Estimated Average Requirement (EAR)
Now that the FNB has spent a decade establishing new RDAs for many nutrients, carefully evaluating the data and increasing the recommendations for some nutrients for some age groups, the labeling committee recommends tossing all that aside and basing nutrition policy and labeling instead on the Estimated Average Requirement (EAR). This is the amount of a nutrient that is supposed to meet the actual nutritional need for half of a specific population. By definition, this amount is below the actual nutritional needs of half that population. Once an EAR has been established, a safety factor is added to it, increasing it by two standard deviations (about 20% for most nutrients). This higher number is the RDA, which is calculated to cover the needs of 97.5% of the population. Why should nutrition labeling and nutrition policy in the U.S. be targeted toward meeting the needs of only half the population, instead of meeting the needs of almost all?
But waitit gets worse. Not only does the committee recommend using the EAR instead of the RDA-it recommends using a population-weighted EAR. So what?
Well, let's look at vitamin C as an example. Vitamin C was one of the nutrients whose RDA was increased in the new DRI reports. The RDA for adults had been 60 mg, but the new reports increased it to 75 mg for women and 90 mg for men-a 50% increase for men. If nutrition labeling were to be based on the highest RDA, the new reference value for vitamin C would be 90 mg. But instead, the IOM's committee suggests using the EAR, which is only 60 mg for women and 75 mg for men. The values are even lower for children and teens, so the population-weighted reference value would remain stuck at or near 60 mg. Thus, under the labeling committee's recommendations, absolutely no public benefit is derived from the FNB's decision to revise the vitamin C recommendations upward. (See Table 1)
Iron provides another example. The RDA for iron is 18 mg for women of childbearing age, 8 mg for adult men and postmenopausal women and somewhere in between for children and teens. The EAR is lower-8 mg for women of childbearing age and between 4 mg and 8 mg for everybody else. The current label reference value for iron is 18 mg, and the new one under the scheme proposed by the IOM's labeling committee would be a mere 6.1 mg-only one-third of the RDA for women of childbearing age and in fact below the EAR for those women. (See Table 1)
The FNB recommended higher intakes of vitamin D for the elderly, but these increases would be lost in the population-averaging approach. Supplemental vitamin B12 was also recommended for the elderly-in an amount that is greater than the new reference value, so a breakfast cereal or a dietary supplement labeled to contain 100% of the DV for vitamin B12 would not, in fact, contain the supplemental amount now recommended for the elderly. This would be a shame, brought about by a wrongheaded approach to nutrition labeling, if FDA were to adopt the labeling committee's recommendations.
The DRI reports all include a basic discussion of the purposes of the RDA as compared to the EAR. The RDA is clearly identified as the target intake for the individual. Thus, for purposes of nutrition education, it is the RDA that should be the focus of communications to the consumer. Nutrition labeling is clearly a major form of communication to the consumer regarding target levels of nutrient intake. The percent DV shown on the label is not an abstract figure. It gives the consumer information about the fraction of the daily need for each nutrient provided by the product. For some products, such as fully fortified breakfast cereals and dietary supplements formulated to provide 100% DV for a variety of nutrients, the label value is not only a tool for consumer education but is also the basis for product formulation.
If the sole purpose of nutrition labeling were to compare one food to another, either the RDA or the EAR would do, and the committee recognizes this. Why then opt for a marginal standard rather than a more generous one? Could it be so that foods "look better" to consumers? Surely this was not the motive, but it would definitely be the outcome. Foods that contain at least 10% of the reference value for a particular nutrient can be called "good sources" of that nutrient, and foods that contain 20% can be called "excellent sources." If the reference standard is lowered, as the committee recommends, then more foods will meet these requirements.
The labeling committee expresses concern about the potential for over-fortification of the food supply, and it appears that one objective is to decrease overall levels of fortification. For example, a breakfast cereal that now provides 100% of the DV for a variety of vitamins and minerals would, in the future, contain less of each nutrient and still be able to claim 100% of the DV. The same would apply to vitamin and mineral supplements. Would people be pleased to know that a bunch of scientists are proposing to reduce the amounts of nutrients in their fortified foods and multivitamins, or would people be outraged at this subterfuge? The latter seems more likely.
This might all be a tempest in a teapot if the RDA had no real significance to the consumer and if replacing it with the EAR were purely a policy decision with little or no real impact on consumer health. However, the DRI reports repeatedly emphasize that an individual should strive to consume RDA levels of vitamins and minerals. The following is a paraphrase of the advice for the individual from the DRI reports (emphasis in original):
If a person's usual dietary intake is at least as high as the RDA, there is little likelihood that the intake will be inadequate.
If a person's usual intake is in between the RDA and the EAR, there is a great deal of uncertainty about whether that intake is inadequate-it will be inadequate for 3-50% of the individuals in that person's age and gender group. If intakes between the RDA and the EAR are maintained over a prolonged time period, a person might demonstrate signs of inadequacy.
If a person's usual intake is less than the EAR, there is a high likelihood that intake is inadequate.
Summary
Since the DRI reports recognize that intakes near the EAR may lead to signs of inadequacy over time, why would any policymaking agency select the EAR as the basis for nutrition labeling and related efforts in nutrition education? The rational choice is to select the RDA as the appropriate standard and to select the highest RDA rather than averaging out the needs of vulnerable groups.
The FDA would be well advised to set the labeling committee's recommendations aside and stick with the tried and true method of choosing a population-coverage approach to nutrition labeling. The highest RDAs have served consumers well as targets for healthy nutrient intakes for 30 years. FDA should continue to use high values as targets for nutrition policy and nutrition labeling for at least 30 more years. This means using RDAs rather than EARs as the basis for the reference value and using the highest value for each nutrient rather than applying a population-weighted average. NW
About the author:
Annette Dickinson, Ph.D., is president, Council for Responsible Nutrition (CRN), Washington, D.C. She can be reached at 202-776-7929; Fax 202-204-7980; E-mail: annette@crnusa.org.