© Whole Foods magazine

May 2001

Preventing Cancer With Selenium: An Update

By Richard A. Passwater, Ph.D.


This special issue of Whole Foods is dedicated to natural approaches for dealing with cancer, and the editors have asked me to bring readers up to date on my research with selenium and cancer prevention. I didn't even have to twist their arms to get the chance. But I did check to see if they were teasing me or not.

Well, believe it or not, it has been quite some time since I have written on this subject, although I have occasionally interjected a comment or two in these columns whenever I could slip one in. The last time I wrote extensively about my research with selenium and cancer prevention was for my 1999 book, All About Selenium, published by Avery/Penguin Putnam, located in New York, NY.

Actually, I really have been "champing at the bit" to get a chance to talk about three topics:

1. The faulty assumptions that astonishingly led to underestimating the new selenium Recommended Dietary Allowance (RDA).

2. The misconception that all forms of selenium have equal effect.

3. The concept that the anti-cancer effect of some selenium compounds is related to their nutritional value or toxicity.

I also will briefly comment on and rank selenium supplements according to their nutritional and cancer prevention potential. Some selenium compounds are very powerful anti-cancer compounds. The published literature describes selenium compounds having 80% efficiency in preventing cancer. However, our studies found selenium compounds that afford even greater protection.

As many of you know, I have been researching this area for more than 40 years, and I am a patent-holder in the field. I have been conducting laboratory animal studies with selenium long before the first report of my research appeared in Prevention magazine in 1971. My first patent application for the synergistic actions of selenium and vitamin C was in 1970, before selenium was accepted as an essential nutrient. This also was my first patent application specifically for the anti-cancer action of selenium, which was based on my earlier research from 1959 through the 1960s. My first article on the subject, "Cancer: New Directions," appeared in American Laboratory in 1972. I described my selenium research in Supernutrition: Megavitamin Revolution (Dial Press) in 1975 and further described my research in the first of my four books on selenium, Selenium As Food & Medicine, in 1980.

My colleague in much of this research over the years has been Dave Olson. We were colleagues in the early 1960s at the research laboratory of the General Chemical Division of Allied Chemical Corporation in Marcus Hook, PA. Dave again joined me in the 1970s at American Gerontological Research Laboratories (AGRL), where I was the director of research. When that company merged with Life Science Laboratories, Inc., Dave became president and continued the laboratory research.

In 1996, Dr. Larry Clark and his colleagues published their large, prospective, randomized, placebo-controlled, double-blind Nutritional Prevention of Cancer (NPC) clinical study in the Journal of the American Medical Association (JAMA). This landmark research effort showed that daily supplementation of diets with 200 mcg of selenium yeast cut the cancer death rate in half. At that point, many selenium researchers thought the committee responsible for setting the RDA for nutrients would raise the RDA for selenium to reflect this fact. We were amazed and shocked to find that the committee, although having a few selenium experts, actually decreased the selenium RDA. Many of us believe that such a decision flies in the face of the available facts. I will explain shortly. (also see Neve, Nutr Rev. 2000)


Let me begin by dealing with the improper assumption that "selenium is selenium is selenium." With deference to Billy Shakespeare, it may be true that "that which we call a rose by any other name would smell as sweet," but one selenium compound is not the same as another. It is a fallacy to lump selenium compounds together and generalize them all as "selenium." Unlike roses, not all selenium compounds have equal biochemical value or similar actions. Yet, nearly all nutritionists group all selenium compounds together as if they were merely atoms of the element selenium that were being bandied about the body without taking into regard the biochemistry of the entire molecules in which the selenium atoms are incorporated.

There are many different selenium-containing compounds that have various biochemical actions. Incidentally, selenium in its elemental (metal) form is not useful to the body, whereas the selenium ions are to a degree. If it has been a while since your last chemistry class, ions are atoms that have an electrical charge because one or more electrons have been removed from or added to them. It is important to know, too, that the form of selenium that is most useful in the body is selenium that is incorporated as an integral part of an organic molecule, such as a selenoprotein.

We don't consider all sulfur compounds as being "sulfur." We distinguish among the sulfur-containing compounds cysteine, methionine, MSM, SAMe, biotin, lipoic acid, N-acetyl cysteine (NAC), etc. and speak of the individual needs for the dietary essential amino acid methionine and the dietary essential vitamin biotin, etc. Nor do we consider all nitrogen compounds as being "nitrogen." We recognize the distinctions among the nitrogen-containing compounds lysine, nitric oxide and spermine. Moreover, we speak of the individual actions of the dietary essential amino acid lysine, the multifunctional radical nitric oxide, and the polyamine growth factor spermine. We don't consider all carbon compounds as just being "carbon." We see differences among glucose, linoleic acid and amylase and speak of the individual needs for maintaining glucose (blood sugar) levels, obtaining adequate amounts of the dietary essential fatty acid linoleic acid, and secreting amylase to digest starches.

Elementary, yes! Then why are all selenium compounds usually lumped together as "selenium?" This has been a frustration to me for about 40 years. Nonetheless, it does tip me off as to just how little an "expert" knows when he or she speaks in terms of "selenium" in general without qualification as to specific selenium compounds.

Scientists should keep in mind that selenium compounds vary in their suitability for nourishing the body. Different selenium compounds are metabolized in various ways. They can affect the metabolic pool of selenium compounds and metabolic pathways differently. Different selenium compounds have diverse functions in the body. And, importantly, different selenium compounds vary in their cancer-protection and even cancer-destroying abilities. Some selenium compounds have very little, if any, anti-cancer properties, while others are very powerful anti-cancer compounds.

Added to this complexity is the fact that the body conserves selenium compounds very carefully and has a hierarchy as to which selenium compounds are made from the pool of selenium ions or compounds that is available. Further, the body has a "pecking order" that determines which organs are first to receive these selenium compounds. (Neve, Nutr Rev. 2000) The body regulates this hierarchy via upregulation and down-regulation of the genes responsible for making the various selenium compounds and their transporters and receptors. Only when there is adequate selenium in the selenium metabolic pool are all of the desired selenium compounds made in the body and circulated to all organs in abundance.

The vast majority of people do not have optimum selenium reserves in the metabolic pool. Please keep in mind that the NPC clinical trial found that people having normal diets who also took selenium supplements had half the cancer death rate as those with normal diets and not taking selenium supplements. If the average American is getting enough selenium as the new RDA would suggest, then why would taking 200 micrograms more of selenium each day cut the cancer death rate in half?


Why do we need selenium? Wouldn't sulfur do the job? What does selenium bring to the table that other elements don't? Couldn't nature just make other sulfur-containing compounds do the job without resorting to using another element? No! The atoms of the element selenium provide nature with a neat tool to build molecules that are powerful and versatile reducing agents. An oxygen free radical can react with a lipid (fat) molecule to form a lipid peroxide which, like hydrogen peroxide, is not a free radical but a reactive oxygen species (ROS) that can cause harm to body components, especially cell membranes.

Some selenium compounds can stop some ROS chain reactions and even repair free radical damage. For example, the family of selenium-containing compounds called glutathione peroxidases repair phospholipid peroxide damage in membranes, thus breaking the peroxidation chain reactions that damage cells. Also, another family of selenium-containing compounds repair epoxide damage in DNA, thus repairing abnormal DNA and preventing mutations that can lead to cancer. Meanwhile, various selenium compounds have several other biochemical functions as well.

When I first began studying these then unidentified selenium compounds, I was not sure if they were one and the same compound or different selenium compounds. I followed selenium biochemistry by monitoring the biological functions as if multiple compounds were responsible until it could be shown that the various functions were due to one and the same compound. Since I did not know the protein structure or structures involved, I studied them as "selenium factors." Unidentified "black box compounds," which I called "selenium peroxide reducing factor," are now known to be a family of five glutathione peroxidases. Furthermore, what is now known as a large family of cellular enzymes I termed "selenium epoxide reducing factor."

The following paragraph is not essential to your understanding of why selenium is unique and especially beneficial for our bodies, so you may skip it if you choose. But if you are scientifically inclined, you will want to read it.

On the periodic chart, selenium is found in group 16 (formerly known as group VIA) just beneath sulfur. This means that both sulfur and selenium atoms have outer electronic shells that contain six electrons. Thus, their chemical reactions are very similar. Selenium, however, is in the next lower period and is a metalloid, whereas sulfur is a nonmetal. As a metalloid, selenium can behave as a metal by donating electrons during a chemical reaction and behave as a nonmetal by accepting electrons. Selenium atoms have a ground state electronic shell distribution of [Ar]3d(10)4s(2)4p(4) compared to sulfur's [Ne]3s(2)3p(4). This electronic shell structure gives selenium the versatility to readily accept or donate electrons and makes the selenium atom an ideal catalytic center and a great semiconductor. More important for us, the selenium atoms in organic selenium-containing compounds can function as redox centers.

As I mentioned, the same number of electrons in the outer electronic shells of selenium and sulfur means that they can have similar chemical reactions. Plus, they have similar atomic sizes, bond energies, ionization potentials, electron affinities, and electronegativities. This explains why selenium is sometimes mistakenly incorporated into proteins by plants and animals.

Still, nature needs selenium for reasons that sulfur can't fulfill. Please keep in mind that selenium is the only trace element that our genes specify to form an amino acid that is then incorporated into proteins. The codon UGA species selenocysteine, the 21st amino acid. Selenocysteine is not an accident and is the keystone compound needed to make several selenium-containing proteins (selenoproteins). In humans, it does not appear to any significant extent that selenocysteine is non-specifically incorporated into proteins by mistake instead of cysteine. Although selenocysteine is not dietary essential, having more in the diet assures the body of having a more efficient supply available to make these critical selenoproteins. This is why I chose selenocysteine as my primary organic selenium source in my laboratory animal experiments conducted during the 1960s and very early 1970s.

The selenium that most often gets into proteins by accident is in the form of selenomethionine, which can readily be incorporated by accident in place of methionine. Selenocysteine and other selenoamino acids are formed this way, too, but to a lesser degree. The difference between methionine and selenomethionine is that the methionine molecule contains an atom of sulfur whereas selenomethionine molecules have an atom of selenium instead. Plants make selenomethionine in ratio to the amount of selenium in the soil. Soil selenium varies widely from region to region, and thus, a given plant will have differing amounts of selenomethionine according to where it grows. Food values for selenium content are meaningless, as the content is so variable.

It was once contended that selenomethionine was used to make so-called selenomethionine-specific proteins including thiolase, beta-galactosidase and certain muscle proteins in humans. (Sunde, Annu. Rev. Nutr. 1990) The arguments are not convincing that these proteins are indeed specific human selenoproteins by design. Rather, they appear merely to be inadvertent selenium-containing proteins formed randomly by accident. True, these selenomethionine-containing proteins often outperform their sulfur-containing analogs, but since no one has yet found a gene that specifies for a selenomethionine-containing protein, most of us today remain skeptical.

Plant-formed selenomethionine enters the food chain directly, in plant-based foods, or via animals that have eaten the plants. Humans and other animals cannot make selenomethionine directly from inorganic selenium. (Cummins & Martin, Biochem. 1967) Our bodies' metabolic pool of selenomethionine depends on dietary selenomethionine and the recovery of selenomethionine from proteins. Since our genes don't appear to specify the production of selenomethionine, and it appears that little is obtained via conversion of selenocysteine (as is the case with their sulfur analogs, which can readily convert back and forth between methionine and cysteine), then selenomethionine would have to be dietary essential, which I don't believe is the case. Incidentally, although selenomethionine cannot readily be made from selenocysteine in humans, selenocysteine is readily made from selenomethionine.

Until recently, nutritionists thought that there were only 20 amino acids that the human body uses to make proteins. Of these 20 amino acids, only 10 are dietary essential. The other 10 are essential to life, but they do not have to be in the diet as they can be made in the body from the 10 dietary essential amino acids. Thanks to the fact that our DNA specifies for selenocysteine production, selenocysteine does not have to be in the diet, as long as there are selenium ions available in the metabolic pool. About 80% of the body's selenium may be present as selenocysteine. (Hawker et al., J. Inorg. Biochem. 1985).


The need for "selenium" is not merely a question of how much "selenium" you need to eat, but how much biochemically useful selenium you need to eat. What is important is which selenium compounds end up in the tissues and circulate in the blood, and how much of these compounds is present. The selenium compounds that circulate in the blood are of three types-1. selenoproteins and other selenium-containing organic compounds; 2. organic compounds that inadvertently contain selenium; and 3. inorganic selenium compounds.

In the past, the selenium RDA was set with regard to the first known selenoprotein, glutathione peroxidase, in mind. Now we know that there are at least five glutathione peroxidases and apparently each has a different selenium requirement. As Dr. Jean Neve of the Free University of Brussels states, "This makes it more and more questionable to base nutrition recommendations for selenium on a single biochemical chemical function .... Conclusions drawn from the response of one particular selenoprotein do not apply to all biological functions demonstrated to depend on the element." (Neve, Nutr Rev. 2000)

The new RDA Committee did make an improvement over the "single selenium compound tells all we need to know concept." It chose two compounds to monitor. A main assumption used in setting the new selenium RDA is that the nutritional need for selenium is related to optimizing the blood levels of only two selenium-containing proteins, glutathione peroxidase-3 and selenoprotein P. The committee correctly reasoned that the dietary selenium that ends up inadvertently incorporated into proteins in which the selenium serves no purpose should not be considered, as it serves no function.

Nevertheless, the reasoning of the Committee still is faulty. Its first fallacy is in assuming that the two selenoproteins that it monitored in the bloodstream are the only selenoproteins of importance for body function and/or cancer prevention. The Committee members studied the effect of dietary selenium by noting the effect of increasing selenium intake on the blood levels of only these two selenoproteins. When the blood levels of these two selenoproteins no longer increased with the amount of selenium in the diet, they assumed that this was the RDA. Meanwhile, there are many more selenoproteins that were not measured. Thus, the Committee was unaware of the continued rise in blood levels of these selenoproteins with the higher amounts of dietary selenium. Improved analytical procedures show that there are at least 35 selenoproteins (Behne et al., l0th International Symposium on Trace Elements in Man and Animals, 2000). That there are many more is suggested by a recent study, which indicates that there are eight selenoproteins in artery walls, eight selenoproteins in brain tissue and nine selenoproteins in testis. (Qu et al., Biolog. Trace Element res. 2000) I was very disappointed that the Committee did not even consider my favorite selenoproteins, the thioredoxin reductases, the glutathione S-transferases and their mimics or even selenoprotein W.

The second fallacy by the RDA Committee is in ignoring so much of the published scientific literature about selenium. The group narrowly defined the studies that qualified for its consideration, ultimately settling on just two. One was a Chinese study and the other a New Zealand Study. In the Chinese study, individuals were of lower body weight and the level produced only two-thirds of the full expression of glutathione peroxidase.

The Committee discarded the NPC study because it had not yet been replicated. Please remember that the NPC clinical trial found that people cur not normal diets who also took selenium supplements had half the cancer death rate as those with normal diets who did not take selenium supplements. We ask again, "If the average American is getting enough selenium as the new RDA would suggest, then why would taking 200 mcg more of selenium each day cut the cancer death rate in half?"

The third problem may be in a possible "misinterpretation" of the remaining study. From the view of the Committee, it applied different statistical regression analysis procedures to the study and obtained a "more conservative" conclusion differing from that of the original New Zealand scientists. (Duffield et al., Amer. J. Clin. Nutr 1999). Using the same data, Dr. Margaret Rayman of the University of Surrey (England) estimates the RDA to be at least what was suggested previously. (Rayman, Lancet 2000)

Considering these three seemingly forced or even contrived limitations, the new selenium RDA is not on very sound footing.



To date, the only selenium supplement that has been studied for its anticancer action in large clinical trials is selenium yeast. Selenium yeast was used in both the Chinese study and the NPC study. As Dr. Larry Clark revealed when I interviewed him for this column in February 1997 and again in September 1998, he decided to use selenium yeast because it has a measured level of organically bound selenium compounds. Nevertheless, selenomethionine and inorganic selenite also have been widely used in dietary supplements.

Nearly all selenium compounds contribute to both the body's nutritional needs for selenium and its defense against cancer. However, the various selenium-containing compounds may work through several different mechanisms. Various selenium compounds have been shown to inhibit cancer in various organs, including the liver, skin, pancreas, colon and the mammary gland. The key appears to be to introduce the selenium in a dose-dependent manner (increasing amounts of selenium compounds results in increasing cancer protection) until toxic limit is reached.

When tests were performed, certain selenium compounds performed less well than others, for example, trimethylselenonium was essentially ineffective. Certain alkyl selenocyanates had only moderate anti-cancer effects, with the greater effect going to the longer alkyl chain homologs Among the selenides, methylphenyl selenide was eight times more effective than diphenyl selenide. Selenium-methylselenocysteine proved to be an effective compound. Triphenylselenonium was fairly effective. Long ago, Dave Olson and I conducted research showing that the most powerful anti-cancer compounds are the allyl phosphine selenides. Dr. Clemet Ip of The Roswell Park Cancer Institute in Buffalo noted, “Small changes in the structure of selenium compounds can lead to rather surprising changes in biological activity.” (Ip, J. Nutr. 1998) Yet the RDA experts still speak of selenium as if the only concern were the overall amounts of elemental selenium present. As Yul Brynner said in The King and I, “Is a puzzlement!”




Several epidemiological (population) studies show that people with higher levels of selenium compounds in their blood have less cancer than those having lower selenium levels. Dr. Neve points to "accumulating evidence [that] suggests that selenium has further beneficial effects at doses higher than those regarded as adequate based on previous criteria." (Neve, Nutr Rev, 2000) Dr. Rayman states, "Even selenium replete individuals find further enhancement of their immune system with higher selenium intake." (Rayman, 15t International Conference on Human Functioning, 2000)

This is not to say, however, that selenium compounds do not provide anticancer protection until the so-called nutritional needs are well-supplied. Some of the standard nutritional functions of selenium include the formation of selenoproteins such as the glutathione peroxidases which protect cell membranes against free radical damage. This is just one example of a selenium anti-cancer mechanism. At low selenium intake levels, the body cuts back on glutathione peroxidase production to help make selenium available for conversion into other selenoproteins.

Other selenoproteins are involved in immune system enhancements and this too is an anti-cancer action. Selenium compounds bring several anti-cancer mechanisms to the table. Some of these mechanisms involve prevention (membrane protection, DNA protection, etc.) and others (epoxide repair, immune stimulation, apoptosis, etc.) involve cancer cell destruction.

Selenium compounds do not have to enter the selenium metabolic pathway to produce their anti-cancer effect. (Ip, J. Nutr. 1998). The anti-cancer effect of selenium compounds is not related to formation of selenoproteins. (Ip and Ganther, Cancer Res.1996) The anti-cancer effects of selenium compounds are not related to the tissue accumulation of the compounds. (Ip et al., Anti-cancer Res.1998)

As one's selenium intake is increased into the optimal range (which, obviously, is below the toxic range), the number of different selenium compounds that provide anti-cancer action increases. Again, please keep in mind the fact that the NPC clinical trial found that people with normal diets who also took selenium supplements had half the cancer death rate as those with normal diets who were not taking selenium supplements. If the average American is getting enough selenium as the new RDA would suggest, then why would taking 200 mcg more of selenium each day cut the cancer death rate in half?

If one takes specific selenium anticancer compounds that don't have to enter the body's selenium metabolic pool, then greater anti-cancer action can be achieved independent of other selenium compounds in the diet. And, if these selenium compounds are powerful enough, greater anticancer action can result from even relatively low levels of selenium in the body. The contention that critics unskilled in selenium biochemistry often present is that any cancer protection comes only from toxic levels of selenium. They wrongly argue that this precludes selenium as a viable candidate for chemoprevention. Nonsense! Selenium's anti-cancer action is independent of its nutritional role and toxicity. The very powerful and safe selenium anti-cancer compounds that were developed by Olson and myself decades ago and further studied by Olson, are independent of selenium's nutritional status.


Presently, there are three basic types of selenium supplements: high-selenium (selenized) yeast, selenomethionine, and inorganic selenium. Better selenium supplements are expected soon.

Of least value is inorganic selenium. However, even inorganic selenium compounds can increase the body's metabolic pool of selenium ions and thus contribute to the production of selenoproteins that have anticancer actions. However, the production of selenium anti-cancer compounds from inorganic selenium compounds (chiefly selenates and selenites) is inefficient. The inorganic selenium first has to be converted into compounds that can enter the selenium metabolic pool and then be converted into specific organic anti-cancer selenium compounds.

Several researchers have successfully used inorganic selenium in their laboratory animal experiments mainly because inorganic selenium is cheap and easy to use. For example, it can be added to the animals' drinking water. By contrast, anyone working with organic selenium compounds would have to blend these substances into the lab chow. Inorganic selenium produces rapid short-term increases in some beneficial organic selenium compounds, but these increases are relatively short-term. However, the results obtained with inorganic selenium are significant, but not spectacular.

Very little inorganic selenium enters the human food supply. This is good because toxic effects from intake of inorganic selenium occur at quantities far less than those from organic selenium compounds. One laboratory animal study links inorganic selenium compounds at moderate concentrations to eye cataracts.

Selenomethionine is next in preference as it is a form of selenium normally present in the human food supply. Selenomethionine also can provide selenium for nutritional needs and for anticancer actions. Selenomethionine readily can increase the body's selenium metabolic pool, but much of the selenomethionine intake can be temporarily incorporated nonspecifically into proteins in place of methionine, and not be available to the selenium pool again until that protein is degraded by the body. Thus, the benefits of selenomethionine occur over a longer time frame and do not produce rapid results.

Selenomethionine can be converted into selenocysteine readily, and the latter is the building block for the selenoproteins. Selenomethionine is significantly less toxic than inorganic selenium.

The best-studied selenium supplement in human clinical trials so far is high selenium yeast (also called selenized yeast). As mentioned earlier, this was chosen for the NPC study because it is a cocktail of organic selenium compounds. It once was thought that selenium yeast contained about 50% of its selenium compounds in the form of selenomethionine. More sophisticated analyses now suggest that selenomethionine accounts for no more than about 20% of all selenium-containing compounds. Another group of selenium-containing compounds includes selenocysteine, seleno-methyl selenocysteine and selenomethionine. There are other groups of compounds mostly unidentified at this time that make up about 40% to 50% of all the selenium compounds in yeast. These unidentified compounds may include selenophosphates, triphenyl phosphine selenide, diselenides, triselenides and other very interesting organic selenium compounds.

Selenium yeast provides compounds that are direct anti-cancer compounds not having to go through the selenium metabolic pool. The effects of selenium yeast are both immediate and long term.

Of these three selenium supplements, the advantage goes to selenium yeast. However, more effective powerful selenium anti-cancer supplements and drugs are on the way. Stay tuned. WF

© 2001 Whole Foods Magazine and Richard A. Passwater, Ph.D.

This article is copyrighted and may not be reproduced in any form (including electronic) without the written permission of the copyright owners.