RDAs And Safe Upper Levels: Solid Science Versus Bureaucratic Bias
5 March 2001
by Wyn Snow, Managing Editor

Recently, the national Institute of Medicine (IoM) has cautioned that overdosing on vitamin A can cause birth defects and liver damage. Down similar lines, a letter published in the British journal Nature warns that taking 500 mg of vitamin C daily might be harmful. Other stories about potential dangers of excessive vitamin and mineral intake appear regularly in the media.

Yet other nutrition experts and numerous supplement producers say most supplements are safe. Indeed, Dr. Adrianne Bendich (coeditor of three books and two Annals of the New York Academy of Sciences and chapters in other technical books, as well as author and coauthor of more than 75 scientific articles on vitamin biochemistry and safety) performed in-depth analyses of more than 50 years of literature on the safety of five vitamins (including vitamins A and C) -- and notes in an interview in Whole Foods Magazine, "It is remarkable to find that almost all of the safety issues often mentioned are not based on solid data."

At the other end of the scale, the Institute of Medicine (IoM) also says most Americans get enough vitamins and minerals from their food (with a few exceptions such as calcium and folate) -- yet nutrition experts told participants at a conference held at the Children's Hospital Oakland Research Institute last June 2000 that many Americans with the lowest income do not get enough vitamins. Bruce Ames of the University of California at Berkeley says, "Wherever you turn around . . . the poor are eating such poor diets I think they are battering their DNA, causing cancer and maybe damaging their brains."

With such disparate opinions being voiced by experts, how do consumers decide what to do? Should we simply eat more fruits and vegetables? Take a multivitamin "just in case" our diets don't provide enough? And what about those "mega" doses of vitamins and minerals that some people take to optimize their health? With almost half of all Americans taking supplements of one kind or another, such questions pose serious public health concerns.

New guidelines from Institute of Medicine

The national Institute of Medicine (IoM) recently issued new dietary guidelines -- covering Recommended Dietary Allowances (RDAs) and safe upper levels -- for two vitamins and several minerals. It is the fourth in a series of reports beginning in 1997 that have changed the RDA landscape into an alphabet soup of "dietary reference intakes." If EARs, RDAs, AIs and ULs are unfamiliar and confusing, you are not alone when it comes to understanding nutritional terminology (see sidebar explaining these terms).

What do these new dietary guidelines mean? Are they reliable, objective, useful and comprehensive? Are they solid science or partisan social policy? Reconciling complex and sometimes contradictory "standards" can be perplexing and even a health hazard. Just as importantly, non-critical acceptance of supposedly objective science can have far-reaching consequences in terms of health policy and legislation.

Recently, SupplementQuality.com asked several experts for comments on these issues while placing a magnifying glass on some of the fine print in four lengthy reports issued to date. Some of our findings are both surprising and controversial. We discussed our concerns with Professor Robert Reynolds, a world-renowned researcher at the University of Illinois at Chicago, who says, "RDAs and safe upper levels are more political than scientific." (See our interview with Reynolds.)

Still plenty of guesswork

You might think that RDAs represent clearly defined values, derived from extensive scientific data, that say how much of a nutrient you need to maintain optimal health. But our source at the Institute of Medicine (IoM), acknowledged that "There's still a lot of guesswork involved."

RDAs are actually estimates, based on surprisingly limited studies, often involving small numbers of people and incomplete data. Essentially, the IoM does the best it can with sketchy evidence to establish either a recommended dietary allowance (RDA) or an adequate intake (AI) level.

Just as identifying a fingerprint depends on being able to see enough of the lines and whorls, establishing an accurate RDA depends on having enough data to extrapolate from a small number of people to the overall population. In assessing the evidence, we reviewed parts of several lengthy IoM reports in an attempt to judge the completeness of the evidence.

Just how complete is the evidence?

We began with thiamin (vitamin B1), the first vitamin to be isolated in pure form (1926), thinking there would be ample underlying research. We discovered that the RDAs for thiamin are based on 9 studies of 105 people. Surprised to find what we thought was a rather small number of people, we contacted the IoM to see if these limited studies were accurate and typical.

Paula Trumbo, Senior Program Officer of IoM's Food & Nutrition Board, says, "Some of the RDAs are based on smaller samples. However, research indicates that for the most part, human metabolism varies within a fairly narrow range -- so that 97% of the population falls within 20% of the peak of the bell-curve distribution. Most vitamins are enzyme cofactors, so the body's use of them is closely tied to metabolism." (See sidebar for explanation of enzyme cofactors, vitamins, minerals, and metabolism.)

Thus, the IoM first determines if the data provide enough detail about a nutrient to determine an estimated average requirement (EAR): the daily intake that meets the nutritional needs of 50% of the people involved in the study and presumably in the population at large. If so, this establishes a central peak for the bell-curve distribution -- and statistical methods can then be used to calculate an RDA. But if an EAR cannot be determined from the data for a nutrient, the IoM develops an adequate intake (AI) level, which is also an estimate. (See sidebar for explanation of bell-curve distribution and how the RDA and AI are calculated).

One persistent difficulty the IoM faces is that data may be plentiful for one age group or gender, but sparse for another. In other words, as you may have observed yourself, it's important to adjust intake levels for age, gender and activity factors, as well as individual issues. The IoM, faced with such unknowns, extrapolates -- using data from one gender or age group as the basis for the RDA or AI for another. In the reports we reviewed, the chapter for each nutrient ends with a section describing additional research the IoM feels would be valuable. A significant fraction of their recommendations deal with filling the gaps in currently available data. (See sidebar's excerpts from the IoM reports for calcium and vitamins C, D, and E.)

Minimalist RDA design

Here's another thing about RDAs and AIs you may have missed: they're designed to prevent classical deficiency diseases rather than to optimize health. The original RDAs were developed in the 1940s by the military as a dietary guideline for US combatants in World War II. In the years since, they have been expanded to address the needs of women, infants, children and older people as well as young men; responsibility for establishing RDAs moved from the military to the National Academy of Sciences and its Institute of Medicine (IoM). From the beginning, however, RDAs were designed quite literally as the smallest amount of a nutrient that will prevent 97% of the population from becoming ill with scurvy (vitamin C), beriberi (vitamin B1), rickets (vitamin D), and other identified deficiency diseases.

Another thing few Americans realize: the new RDAs won't appear on supplement labels until the IoM completes its fifth report, expected later this year. Until then, labels still use information from the previous set of RDAs published in 1989.

The IoM's new RDAs are beginning to address wider health concerns: for example, the criterion used to establish an EAR for vitamin C is related to its antioxidant properties, and the new RDA is higher than the previous one (now 75 mg for women, 90 mg for men). However, the IoM chose criteria for vitamin E that are "based largely on induced vitamin E deficiencies." Even though the IoM says, "A large and growing body of experimental evidence suggests that high intakes of vitamin E may lower the risk of some chronic diseases," it then characterizes this research as "limited and discordant" and therefore not solid evidence for recommending higher vitamin E intakes.

The IoM takes a conservative stance in deciding whether new research has conclusively proven the value of a nutrient in relation to a disease. However, two other factors also contribute to the IoM's tendency to keep RDAs low.

Does IoM have it right?

In developing an RDA, IoM committee members use a traditional peer review process. First, they decide which criteria to use as a measure of adequate intake -- such as blood levels or urine excretion of a vitamin compound. Next, they select the scientific literature relevant to that criterion and weigh the studies based on the usefulness and relevance of the experimental design. Finally, they discuss their disagreements and reach consensus based on the evidence. However, a closer scrutiny of the IoM's membership and interests uncovers two important sources of potential bias.

Dr. Jonathan Wright, a well known alternative-health practitioner and author whose past conflicts with the FDA have made him a controversial figure, notes an obvious, but little commented-on fact: members of the IoM all espouse a traditional point of view. "They're all coming from the same side of the fence," Dr. Wright says. "No one is looking at what might be called the natural medicine point of view."

Dr. Wright also states the IoM gives little weight to research that doesn't agree with their preconceptions and gives this example:

"Two years ago in the American Journal of Clinical Nutrition (which is regarded as the leading publication in the field), a research team reviewed the safe upper levels of vitamin D. The team's analysis looked first at the view that our bodies have evolved to take a full day's worth of sunshine, then examined the scientific literature. Both supported the notion that the safe upper level of vitamin D is 10,000 IU, making 2,000 IU five-fold too low. The Journal did not present any editorial opposition to this notion, which they often do when they disagree with an author's conclusions. Yet, where has the Institute of Medicine set the safe upper level? It's so far out of touch."

This month's issue (volume 73) of the American Journal of Clinical Nutrition contains another article about vitamin D by these same authors describing their research with 61 people. Vieth, Chan and MacFarlane state that a 100 mcg (4,000 IU) daily dosage of vitamin D3 "effectively increases blood levels of [vitamin D] to high normal concentrations in practically all adults, and [serum levels] remained within the physiological range; therefore we consider 100 mcg to be a safe intake."

In comparison, the IoM chose serum calcium rather than serum vitamin D as their criterion for setting a safe upper level (UL). They set the UL at 50 mcg (2,000 IU), even though they viewed 60 mcg as the largest dosage having no negative side effects and 95 mcg as the smallest dose in their literature that had caused a rise in serum calcium above the level they characterized as normal.

(More information about safe upper levels and how they are calculated appears later in this article.)

Research funding: Another bias?

Dr. Julian Whitaker, author of eight books on nutrition and health, points out that the packaged food industry has a vested interest in keeping RDAs low. This same industry funds much of the nutrition research in the US and more than a few academic professorships, a fact that Whitaker suggests further clouds research objectivity.

According to Dr. Whitaker, "When RDAs are low, it's easy for a juice manufacturer to put 60 mg of vitamin C into a glassful of their product and claim it contains 100 percent of the daily requirement. But if RDAs are set at higher levels, such as 500 or 1000 mg of vitamin C, amounts that many people take to prevent catching cold, then 60 mg of vitamin C in that glass would be a mere 12 or even 6 percent of the RDA, which is a much weaker claim."

From Wright and Whitaker's point of view, the composition of the IoM could be improved by including more points of view among its members, and its conclusions would be stronger if nutrition research was funded by more objective sources. Much university research is funded by food companies, a topic that deserves review in its own right. A third concern deals with how we determine the "normal levels" of various nutrients in the body.

How "normal" is normal?

In nutrition -- just as in other scientific research -- the quality of data sources is crucial to the value of the results.

As Dr. Wright points out, the term normal in statistics means the average. He observes, "Normal blood levels were drawn up some years ago by taking blood samples from a group of people and drawing the usual bell-shaped curve. That's great if you're working with a population that's eating enough of the nutrient, such as 100 organic farmers eating their own produce, versus 100 junk food junkies. [Between the two] you'll get vastly different 'normal' levels. The word normal is used differently in statistics where it means average, and in everyday usage where it means healthy."

Indeed, early nutritional studies of RDA needs were flawed by the fact that they used institutionalized populations -- such as incarcerated prisoners and individuals suffering from congenital retardation, whose dietary intakes could best be described as low-cost and limited, and whose metabolic needs may differ in important ways from the wider population.

Continuing high levels of chronic diseases (heart disease, stroke, cancer, diabetes) and excessive intakes of fat and sugar associated with the fast food diets consumed by many Americans combine to suggest that "normal" is a problematic concept at best. We need to be aware that the data characterizing normal levels may be limited, laced with conjecture and experimental flaws. Individual differences -- long accepted in education as a critical factor in tailoring pedagogical methods to the needs of each student -- occur in human physiology as well, and are not addressed by the RDA's "one size fits all" philosophy.

As science, RDAs do not have the robust character and solid underpinnings we might want in order to view them as substantive and reliable. However, RDAs do provide a useful estimate of a minimum intake amount; getting less than this amount probably puts one's health at risk. But what about maximum amounts? How accurate are the IoM's safe upper intake levels?

Safe upper levels - Do no harm!
In establishing a tolerable upper intake level (UL) for a nutrient, the IoM's goal is to ensure that supplement consumers experience no adverse events (negative side effects) from taking . . .