• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

Food

  • Print
  • Share
  • E-mail

Selenium and a Reduced Risk of Site-specific Cancers, FDA-2008-Q-0323

Back to Qualified Health Claims: Letters of Enforcement Discretion

June 19, 2009

Jonathan W. Emord, Esq.
Emord & Associates, P.C.
11808 Wolf Run Lane
Clifton, Virginia  20124

 

RE: Qualified Health Claim Petition – selenium and a reduced risk of site-specific cancers (FDA-2008-Q-0323)

 

This letter responds to a health claim petition received by the Food and Drug Administration (FDA or the agency) on April 24, 2008, which was submitted on behalf of Julian M. Whitaker, M.D., The Coalition to End FDA and FTC Censorship, Durk Pearson and Sandy Shaw, and Youngevity, Inc.  The petition was submitted pursuant to Section 403(r)(5)(D) of the Federal Food, Drug, and Cosmetic Act (the Act) (21 U.S.C. § 343(r)(5)(D)).  The petition requested that the agency allow qualified health claims characterizing the relationship between selenium from dietary supplements and a reduced risk of site-specific cancers. The petition also requested that FDA allow two other qualified health claims about selenium from dietary supplements, one for reduced risk of "certain cancers" and the other for anticarcinogenic effects in the body.  The petition proposed the following claims as model health claims for selenium-containing dietary supplements.

  1. Selenium may reduce the risk of certain cancers.  Scientific evidence supporting this claim is convincing but not yet conclusive.
  2. Selenium may produce anticarcinogenic effects in the body.  Scientific evidence supporting this claim is convincing but not yet conclusive.
  3. Selenium may reduce the risk of prostate cancer.  Scientific evidence supporting this claim is convincing but not yet conclusive.
  4. Selenium may reduce the risk of bladder and urinary tract cancers.  Scientific evidence supporting this claim is convincing but not yet conclusive.
  5. Selenium may reduce the risk of lung and respiratory tract cancers.  Scientific evidence supporting this claim is convincing but not yet conclusive.
  6. Selenium may reduce the risk of colon and digestive tract cancers.  Scientific evidence supporting this claim is convincing but not yet conclusive.
  7. Selenium may reduce the risk of thyroid cancer.  Scientific evidence supporting this claim is convincing but not yet conclusive.
  8. Selenium may reduce the risk of brain cancer.  Scientific evidence supporting this claim is convincing but not yet conclusive.
  9. Selenium may reduce the risk of liver cancer.  Scientific evidence supporting this claim is limited and applies only to hepatitis b virus-induced forms of the disease.
  10. Selenium may reduce the risk of breast cancer.  Scientific evidence supporting this claim is convincing but not yet conclusive.

FDA received an earlier petition dated February 19, 2008, that requested the same qualified health claims as above.  However, in a letter dated March 4, 2008, FDA informed you that it was not able to acknowledge receipt and begin its preliminary review of this earlier petition because the petition was not complete as required under 21 CFR 101.70.  In response, you supplied the needed information in a supplemental submission received by the agency on April 24, 2008.  FDA then acknowledged receipt of the petition in a letter dated May 8, 2008, which initiated FDA's preliminary review of the petition.  FDA filed the petition for comprehensive review on June 6, 2008, and posted the petition on the FDA website for a 60-day comment period, consistent with the agency's guidance for procedures on qualified health claims.[1]  The agency received no comments in response to the petition. 

The agency received three submissions of supplemental information from the petitioner.  Two submissions provided recent publications on selenium and cancer chemoprevention.  The third submission provided an analysis and opinion of a recent publication on the National Cancer Institute (NCI) Selenium and Vitamin E Cancer Prevention Trial (Lippman et al., 2009).  This submission also included a magazine article written by Dr. Richard Passwater on selenium and prostate cancer studies.  FDA considered the information provided in the supplemental submissions from the petitioner in its evaluation of the petition.  The original decision date for this petition was January 19, 2009.  By mutual agreement, the decision date for this petition was extended to June 19, 2009.

In 2002 FDA received a health claim petition submitted pursuant to Section 403(r)(5)(D), on behalf of Wellness Lifestyles, Inc., requesting that FDA authorize health claims for use on selenium dietary supplements on the relationships between selenium and reduced risk of certain cancers, and between selenium and anticarcinogenic effects (see Docket No. FDA-2002-P-0418 (formerly 2002P-0457)).   On February 21, 2003, FDA informed the petitioners of its intent to exercise enforcement discretion for two qualified health claims describing the relationship between selenium and reduced risk of certain cancers, and between selenium and anticarcinogenic effects:

  1. Selenium may reduce the risk of certain cancers.  Some scientific evidence suggests that consumption of selenium may reduce the risk of certain forms of cancer. However, FDA has determined that this evidence is limited and not conclusive.
  2. Selenium may produce anticarcinogenic effects in the body.  Some scientific evidence suggests that consumption of selenium may produce anticarcinogenic effects in the body. However, FDA has determined that this evidence is limited and not conclusive.

On April 28, 2003, FDA issued a decision letter describing its scientific evaluation of these two qualified health claims and explaining the basis for its intent to exercise enforcement discretion for their use in dietary supplement labeling.[2]

As noted in the petition, FDA published a notice in the December 21, 2007 Federal Register (72 FR 72738; Docket No. FDA-2007-N-0152 (formerly 2007N-0464)) (reevaluation notice) announcing the agency's intent to reevaluate the scientific evidence for the existing qualified health claims for selenium and certain cancers (see Docket No. FDA-2002-P-0418 (formerly 2002P-0457)), along with three other health claims.  However, the discussion in this letter is directed solely to your qualified health claim petition and not to the December 21, 2007 reevaluation notice, which will be addressed separately in a future document.

This letter sets forth the conclusions from the agency's review of the proposed qualified health claims concerning selenium from dietary supplements and certain cancers (claim #1), anticarcinogenic effects (claim #2), and the cancers listed by name above (claims #3-10).  FDA has determined that the current scientific evidence supports qualified health claims on selenium-containing dietary supplements concerning the relationship between selenium supplement intake and a reduced risk of bladder cancer, prostate cancer, and thyroid cancer.  Accordingly, this letter discusses the factors that FDA intends to consider in the exercise of its enforcement discretion for qualified health claims with respect to selenium supplements and a reduced risk of bladder cancer, prostate cancer, and thyroid cancer.  This letter also sets forth the basis for FDA's determination that there is no credible evidence supporting a relationship between selenium supplement intake and a reduced risk of urinary tract cancers other than bladder cancer, lung and other respiratory tract cancers, colon and other digestive tract cancers, brain cancer, liver cancer, and breast cancer. Finally, the letter explains the reasons for FDA's conclusion that claims #1 and 2 are misleading because they are overbroad, fail to disclose material information, and are not supported by the scientific evidence the agency reviewed in connection with your petition.

I. Overview of Data and Eligibility for a Qualified Health Claim

A health claim characterizes the relationship between a substance and a disease or health-related condition (21 CFR 101.14(a)(1)).  The substance must be associated with a disease or health-related condition for which the general U.S. population, or an identified U.S. population subgroup is at risk (21 CFR 101.14(b)(1)).  Health claims characterize the relationship between the substance and a reduction in risk of contracting a particular disease.[3]  In a review of a health claim, the agency first identifies the substance and disease or health-related condition that are the subject of the proposed claim and the population to which the claim is targeted.[4]  FDA considers the data and information provided in the petition, in addition to other written data and information available to the agency, to determine whether the data and information could support a relationship between the substance and the disease or health-related condition.[5]  

The agency then separates individual reports of human studies from other types of data and information.  FDA focuses its review on reports of human intervention and observational studies.[6]  In addition to individual reports of human studies, the agency also considers other types of data and information in its review, such as meta-analyses,[7] review articles,[8] and animal and in vitro studies.  These other types of data and information may be useful to assist the agency in understanding the scientific issues about the substance, the disease or health-related condition, or both, but cannot by themselves support a health claim relationship.  Reports that discuss a number of different studies, such as meta-analyses and review articles, do not provide sufficient information on the individual studies reviewed for FDA to determine critical elements, such as the study population characteristics and the composition of the products used.  Similarly, the lack of detailed information on studies summarized in review articles and meta-analyses prevents FDA from determining whether the studies are flawed in critical elements, such as design, conduct of studies, and data analysis.  FDA must be able to review the critical elements of a study to determine whether any scientific conclusions can be drawn from it.  Therefore, FDA uses meta-analyses, review articles, and similar publications[9] to identify reports of additional studies that may be useful to the health claim review and as background about the substance-disease relationship.[10]  If additional studies are identified, the agency evaluates them individually.

FDA uses animal and in vitro studies as background information regarding mechanisms of action that might be involved in any relationship between the substance and the disease.  The physiology of animals is different than that of humans.  In vitro studies are conducted in an artificial environment and cannot account for a multitude of normal physiological processes, such as digestion, absorption, distribution, and metabolism, which affect how humans respond to the consumption of foods and dietary substances (Institute of Medicine [IOM], 2005).  Animal and in vitro studies can be used to generate hypotheses or to explore a mechanism of action but cannot adequately support a relationship between the substance and the disease.

FDA evaluates the individual reports of human studies to determine whether any scientific conclusions can be drawn from each study.  The absence of critical factors, such as a control group or a statistical analysis, means that scientific conclusions cannot be drawn from the study (Spilker et al., 1991; Federal Judicial Center, 2000).  Studies from which FDA cannot draw any scientific conclusions do not support the health claim relationship, and these are eliminated from further review. 

Because health claims involve reducing the risk of a disease in people who do not already have the disease that is the subject of the claim, FDA considers evidence from studies in individuals diagnosed with the disease that is the subject of the health claim only if it is scientifically appropriate to extrapolate to individuals who do not have the disease.  That is, the available scientific evidence must demonstrate that: (1) the mechanism(s) for the mitigation or treatment effects measured in the diseased populations are the same as the mechanism(s) for risk reduction effects in non-diseased populations; and (2) the substance affects these mechanisms in the same way in both diseased people and healthy people.  If such evidence is not available, the agency cannot draw any scientific conclusions from studies that use diseased subjects to evaluate the substance-disease relationship. 

Next, FDA rates the remaining human intervention and observational studies for methodological quality.  This quality rating is based on several criteria related to study design (e.g., use of a placebo control versus a non-placebo controlled group), data collection (e.g., type of dietary assessment method), the quality of the statistical analysis, the type of outcome measured (e.g., disease incidence versus validated surrogate endpoint), and study population characteristics other than relevance to the U.S. population (e.g., selection bias and whether  important  information about the study subjects -- e.g., age, smoker vs. non-smoker -- was gathered and reported).  For example, if the scientific study adequately addressed all or most of the above criteria, it would receive a high methodological quality rating.  Moderate or low quality ratings would be given based on the extent of the deficiencies or uncertainties in the quality criteria.  Studies that are so deficient that scientific conclusions cannot be drawn from them cannot be used to support the health claim relationship, and these are eliminated from further review. 

Finally, FDA evaluates the results of the remaining studies.  The agency then rates the strength of the total body of publicly available evidence.[11]  The agency conducts this rating evaluation by considering the study type (e.g., intervention, prospective cohort, case-control, cross-sectional), the methodological quality rating previously assigned, the number of studies and number of subjects per group, whether the body of scientific evidence supports a health claim relationship for the U.S. population or target subgroup, whether study results supporting the proposed claim have been replicated[12][13] of the total body of evidence.[14]  Based on the totality of the scientific evidence, FDA determines whether such evidence is credible to support a qualified health claim for the substance/disease relationship and, if so, considers what qualifying language should be included to convey the limits on the level of scientific evidence supporting the relationship or to prevent the claim from being misleading in other ways.

A. Substance

A health claim characterizes the relationship between a substance and a disease or health-related condition (21 CFR 101.14(a)(1)).  A substance means a specific food or component of a food, regardless of whether the food  is in conventional form or in the form of a dietary supplement (21 CFR 101.14(a)(2)).  The petition identified selenium as the substance that is the subject of the requested claims.  Selenium is found in many foods (e.g., meats, seafood, cereals, grains, dairy products, fruits and vegetables), with content being highly variable depending on the selenium content of the soil where the animal was raised or the plant was grown (IOM, 2000).  Selenium is also used in dietary supplements.  As a mineral, selenium is a dietary ingredient under section 201(ff)(1)(B) of the Act (21 U.S.C. § 321(ff)(1)(B)).  Therefore, the agency concludes that selenium, the substance identified in the petition, is a component of food and meets the definition of substance in the health claim regulation (21 CFR 101.14(a)(2)).

B. Disease or Health-Related Condition

A disease or health-related condition means damage to an organ, part, structure, or system of the body such that it does not function properly, or a state of health leading to such dysfunctioning (21 CFR 101.14(a)(5)).  The petition identified "certain cancers," prostate cancer, bladder and urinary tract cancers, lung and respiratory tract cancers, colon and digestive tract cancers, thyroid cancer, brain cancer, liver cancer, and breast cancer as the diseases that are the subject of the proposed qualified health claims.  Cancer is also the subject of the petitioner's proposed claim for selenium and the production of anticarcinogenic effects in the body.

Cancer is a constellation of more than 100 different diseases, each characterized by the uncontrolled growth and spread of abnormal cells (American Cancer Society, 2008).  Cancer is categorized into different types based on specific organ and tissue sites (National Cancer Institute).  Cancers at different sites have different risk factors, treatment modalities, and mortality risk (American Cancer Society, 2008).  Both genetic and environmental (including diet) risk factors may affect the risk of different types of cancers.  Risk factors may include a family history of a specific type of cancer, cigarette smoking, alcohol consumption, overweight and obesity, exposure to ultraviolet or ionizing radiation, exposure to cancer-causing chemicals, and dietary factors.  The etiology (origin or cause of the cancer),[15] risk factors, diagnosis, and treatment for each type of cancer are different (Hord and Fenton, 2007; Milner, 2006).  Since each form of cancer is a different disease based on organ and tissue sites, risk factors, treatment options, and mortality risk, FDA's current approach is to evaluate the evidence for each form of cancer individually when responding to a health claim petition.

Accordingly, the agency considered whether the studies supported a risk reduction relationship for selenium supplements and each of the site-specific cancers for which the petition requested a qualified health claim (i.e., prostate cancer, bladder cancer, lung cancer, colon cancer, thyroid cancer, brain cancer, liver cancer, and breast cancer).   Each of these cancers constitutes a disease under 21 CFR 101.14(a)(5).  Claims #4, 5 and 6 specified "bladder and urinary tract", "lung and respiratory tract", and "colon and digestive tract" cancers. While bladder, lung and colon cancers are considered to be site-specific cancers, there are a number of different cancers that fall under the categories of urinary tract, respiratory tract, and digestive tract cancers.  Urinary tract cancers include cancers of the kidneys, ureters, bladder, and urethra; respiratory tract cancers include cancers of the nose, throat, larynx, trachea, bronchi, and lungs; and digestive tract cancers include cancers of the mouth, esophagus, stomach, small and large intestines, rectum, and anus.[16]  Therefore, FDA reviewed the available evidence on each of these individual cancers in its evaluation of claims #4, 5, and 6. 

In claim # 1, the petitioner requested a claim that selenium may reduce the risk of "certain cancers."  Like claims # 4, 5, and 6, "certain cancers" does not refer to a single cancer, but rather to a group of unspecified individual cancers.  As discussed above, each form of cancer is a different disease based on organ and tissue sites, risk factors, treatment options, and mortality risk.  Because each cancer is different, a claim for "certain cancers," which does not specify the individual cancers for which risk reduction has been found, is too broad and general to be accurate.  In determining whether labeling is misleading, the extent to which the labeling fails to reveal material facts must be taken into account.[17]  In a health claim about reducing the risk of a disease, the disease for which risk is reduced is a material fact because it is one of the essential, characterizing elements of the claim.  See 21 C.F.R. 101.14(a)(1) (defining "health claim" as a claim that characterizes the relationship of a substance to a disease or health-related condition). 

Because each form of cancer is a different disease and generalizations about reducing the risk of cancer are not scientifically valid, the claim "selenium may reduce the risk of certain cancers" is incomplete and misleading because it fails to reveal the individual cancer(s) that selenium may have an effect on.  This could lead a consumer to purchase selenium supplements in hopes of preventing a cancer for which the consumer is at particular risk, but for which there is in fact no evidence of risk reduction from selenium intake.  By referring in general terms to "certain cancers," the requested claim language also suggests that cancers at different sites are essentially the same disease and that it is not important to distinguish between them.  Because the requested claim language "Selenium may reduce the risk of certain cancers" is misleading, FDA does not intend to exercise enforcement discretion for selenium supplements bearing the "certain cancers" claim requested in your petition.[18]

The petitioner also requested that the agency allow a qualified health claim for selenium and the production of "anticarcinogenic effects in the body."  The root word "carcin-", from which "carcinogen" and "anticarcinogenic" are derived, means "cancer."[19]  Thus, the disease that is the subject of the requested claim "Selenium may produce anticarcinogenic effects in the body" is cancer.  However, as discussed above, "cancer" is not a single disease; rather, it is a collective term for a large number of individual diseases that differ with respect to risk factors, etiology, methods of diagnosis and treatment, and mortality risk (Hord and Fenton, 2007; Milner, 2006). Because individual cancers are different, a claim for "anticarcinogenic effects" in general, as opposed to a claim about reducing the risk of one or more individual cancers, is overbroad and misleading because it falsely implies that the same substance (in this case, selenium) can protect against all cancers.  Such a claim is also unsupported by the scientific evidence FDA reviewed.  As discussed below in Sections II and III, the intervention and observational studies from which scientific conclusions can be drawn support qualified health claims for only three of the site-specific cancers reviewed. 

Moreover, although health claims describe the role of a food substance in reducing the risk of a disease or health-related condition, the requested claim language, "Selenium may produce anticarcinogenic effects in the body," is not phrased in terms of risk reduction.  If the requested claim is intended to refer only to risk reduction, it is unclear and overbroad.  For example, because "carcinogenic" means "cancer-causing,"[20] the requested claim could be understood as an absolute prevention claim--i.e., a claim that selenium supplements will be 100% effective in blocking the causes of cancer and thus will prevent those who take selenium from developing any form of cancer.  As discussed below in Sections II and III, the evidence is far too weak to support any such claim. 

Although "produc[ing] anticarcinogenic effects" can be interpreted as a risk reduction claim to the extent that it involves reducing the incidence of cancer by combating cancer-causing substances in the human body, it can also be interpreted in other ways.  The term "anticarcinogenic" and its root, "carcinogen," are broad and have been interpreted in different ways by various medical and scientific bodies.  For example, the National Cancer Institute defines "carcinogen" as "any substance that causes cancer" and "anticarcinogenic" as "having to do with preventing or delaying the development of cancer";[21] the American Cancer Society defines "carcinogen" as "any substance that causes cancer or helps cancer grow."[22]   Under the American Cancer Society definition of "carcinogen" as a substance that "helps cancer grow" in addition to causing it, "produc[ing] anticarcinogenic effects" refers not only to preventing or reducing the risk that someone will get cancer, but also to treating or mitigating existing cancers.  A substance that slows or stops the growth of an existing cancer treats or mitigates that cancer.  Thus, under at least one well-respected medical body's definition of "carcinogen," the requested claim language covers treatment and mitigation of cancer, as well as risk reduction.  Claims about treatment or mitigation of disease are classified as drug claims, not health claims.[23] 

Because the requested claim language "Selenium may produce anticarcinogenic effects in the body" is overbroad, misleading, and unsupported by the current scientific evidence, FDA does not intend to exercise enforcement discretion for selenium supplements bearing the "anticarcinogenic effects" claim requested in your petition.   Based on current scientific knowledge about the differences between individual cancers and on the breadth and uncertain meaning of the phrase "anticarcinogenic effects," the agency believes that no qualified health claim based on that phrase would be truthful and non-misleading.[24]  However, as discussed in the remainder of this letter, the agency has concluded that qualified health claims for selenium and three individual cancers (bladder, prostate, and thyroid) are appropriate for enforcement discretion if the claims are worded narrowly to claim only risk reduction.

C. Safety Review

Under 21 CFR 101.14(b)(3)(ii), if the substance is to be consumed at other than decreased dietary levels, the substance must be a food, a food ingredient, or a component of a food ingredient whose use at the levels necessary to justify the claim has been demonstrated by the proponent of the claim, to FDA's satisfaction, to be safe and lawful under the applicable food safety provisions of the Act. 

FDA evaluates whether the substance is "safe and lawful" under the applicable food safety provisions of the Act.  For conventional foods, this evaluation involves considering whether the ingredient that is the source of the substance is generally recognized as safe (GRAS), approved as a food additive, or authorized by a prior sanction issued by FDA (see 21 CFR 101.70(f)). 

Dietary ingredients in dietary supplements, however, are not subject to the food additive    provisions of the Act (see section 201(s)(6) of the Act (21 U.S.C. § 321(s)(6)).  Rather, they are subject to the adulteration provisions in section 402 of the Act (21 U.S.C. § 342) and, if applicable, the new dietary ingredient provisions in section 413 of the Act (21 U.S.C. § 350b), which pertain to dietary ingredients that were not marketed in the United States before October 15, 1994.[25]  The applicable adulteration provisions of the Act require, for example, that the dietary ingredient not present a significant or unreasonable risk of illness or injury under conditions of use recommended or suggested in labeling or, if no conditions of use are suggested or recommended in the labeling, under ordinary conditions of use (section 402(f)(1)(A) of the Act (21 U.S.C. 342(f)(1)(A))).  Further, a dietary supplement must not contain a poisonous or deleterious substance which may render the supplement injurious to health under the conditions of use recommended or suggested in the labeling (section 402(f)(1)(D) of the Act (21 U.S.C. 342(f)(1)(D))).

Selenium is known to be an essential mineral, but it can also be toxic.  Adverse effects reported from high intakes of selenium included selenosis (hair and nail brittleness and loss), gastrointestinal disturbances, skin rash, garlic-breath odor, fatigue, irritability, and nervous system abnormalities (IOM, 2000).  The IOM (2000) recognized the lowest-observed-adverse-effect level (LOAEL) of selenium intake as 900 micrograms per day (µg/day), and the no-observed-adverse-effect level (NOAEL) of selenium intake as 800 µg/day.  Further, the IOM (2000) characterized the adverse health effects observed at the LOAEL as not severe, but likely not readily reversible, and therefore justifying an uncertainty factor of two.  Dividing the NOAEL (800 µg/day) by this uncertainty factor, the IOM concluded that 400 µg/day is the Tolerable Upper Intake Level (UL) of selenium from food and supplements.  The UL is the level likely to pose no risk of adverse health effects in almost all people.  Based on data from the 1998-1994 Third National Health and Nutrition Examination Survey (NHANES III), the IOM also reported the estimated average intake of selenium from foods and dietary supplements to be approximately 100 µg/day.  FDA has established a reference daily intake (RDI) for selenium at 70 µg/day (21 CFR 101.9(c)(8)(iv)).

Daily intake of a dietary supplement containing 400 µg/day selenium plus intake of 100 µg/day selenium from conventional foods would provide a total estimated intake of 500 µg/day of selenium.  Given the IOM's UL and the current estimated intake of selenium from foods, FDA would likely consider a selenium dietary supplement that encourages intakes above the IOM's UL of 400 µg/day selenium (e.g., through directions for use in the product's labeling or by providing more than the UL in a single caplet or other dosage form) to be misbranded under section 403(a) of the Act.  Such labeling would likely be misleading under section 201(n) of the Act with respect to consequences which may result from the use of the supplement.  Further, a selenium-containing dietary supplement that bears a qualified health claim and encourages selenium intake above 400 µg/day would likely be subject to regulatory action as a misbranded food under section 403(r)(1)(B) of the Act (21 U.S.C. 343(r)(1)(B)), a misbranded drug under section 502(f)(1) of the Act (21 U.S.C. 352(f)(1)), and as an unapproved new drug under section 505(a) of the Act (21 U.S.C. 355(a)).

One form of selenium, selenium sulfide, is reasonably anticipated to be a human carcinogen.[26] As such, the use of selenium sulfide as a dietary ingredient in dietary supplements cannot be considered safe and lawful.  Therefore, dietary supplements containing this form of selenium are not eligible for a qualified health claim.

Section IV discusses the level of selenium that FDA considers necessary to justify the use of a qualified health claim and concludes that use of the qualified health claims described in Sections IV-VI is justified when a dietary supplement contains a "high" level of selenium as defined in 21 CFR 101.54(b).  Under that regulation, a food is "high" in a nutrient if it contains 20 percent or more of the Daily Value for the nutrient.  Thus, based on the current Daily Value for selenium in 21 CFR 101.9(c)(8)(iv), which is 70 µg/day, dietary supplements that contain 14 µg or more selenium per reference amount customarily consumed are "high" in selenium.

FDA concludes, under the preliminary requirements of 21 CFR 101.14(b)(3)(ii), that with the exception of selenium sulfide, the use of selenium in dietary supplements at levels that are "high" as defined in 21 CFR 101.54(b), but no greater than 400 µg/day, is safe and lawful.  

II. The Agency's Consideration of a Qualified Health Claim

FDA has identified the following endpoints, including surrogate endpoints of cancer risk, to use in identifying risk reduction for purposes of a health claim evaluation involving cancer: cases of the site-specific cancer being studied, and recurrent adenomatous[27] colorectal polyps[28] for colorectal[29] cancer.  Adenomatous colorectal polyp recurrence has been used by the National Cancer Institute as a surrogate endpoint for colorectal cancer risk (Schatzkin et al., 1994).  To evaluate the potential effects of selenium on colorectal cancer risk, FDA considered this endpoint as an indicator or predictor of disease.

The petition cited a total of 156 publications as evidence to substantiate the relationships for the claims requested in the petition (see Docket No. FDA-2008-Q-0323).  Several of these publications were provided in duplicate copies.  These publications consisted of 1 research article written in a foreign language; 12 animal studies; 11 in vitro studies; 3 news releases, letters or editorials; 6 government documents; 38 book chapters, review articles, or meta-analyses; and 17 studies that provided no data on the relationship between selenium intake and risk of cancers (Appendix 1).  In addition, the petition provided 16 human intervention and 50 observational studies that evaluated the relationship between selenium intake and one or more cancers.  Lastly, the petitioner submitted two review articles as comments to this petition (Appendix 1). 

In addition to the publications cited in the petition, the public comment period to the reevaluation notice identified 7 intervention and 2 observational studies that evaluated the relationship between selenium intake and cancer risk.  Submissions during the public comment period to the reevaluation notice also identified 1 news release, 5 book chapters, review articles, or meta-analyses, and 2 articles that provided no data on the relationship between selenium intake and cancer risk (Appendix 1) (see Docket No. FDA-2007-N-0464).  In addition to the above, FDA identified 7 intervention and 53 observational studies that evaluated the relationship between selenium intake and risk of site-specific cancers.

A.  Assessment of Review Articles, Meta-analyses, and Book Chapters 

Although useful for background information, review articles, meta-analyses, and book chapters do not contain sufficient information on the individual studies reviewed and, therefore, FDA could not draw any scientific conclusions from this information.  For example, FDA could not determine factors such as the study population characteristics or the composition of the products used (e.g., food, dietary supplement).  Similarly, the lack of detailed information on studies summarized in the review articles, meta-analyses, and book chapters prevented FDA from determining whether the studies were flawed in critical elements such as design, conduct of studies, and data analysis.  FDA must be able to review the critical elements of a study to determine whether any scientific conclusions can be drawn from it.  As a result, the review articles, meta-analyses, and book chapters submitted with the petition or during the public comment period did not provide information from which scientific conclusions can be drawn regarding the substance-disease relationships claimed by the petitioner.

B. Assessment of Animal and In Vitro Studies

FDA uses animal and in vitro studies as background information regarding mechanisms of action that might be involved in any relationship between the substance and the disease, and they can also be used to generate hypotheses or to explore a mechanism of action, but they cannot adequately support a relationship between the substance and the disease in humans.  FDA did not consider the animal or in vitro studies submitted with the petition as providing any supportive information about the substance-disease relationship because such studies cannot mimic the normal human physiology that may be involved in the risk reduction of any type of cancer, nor can the studies mimic the human body's response to the consumption of selenium.  Therefore, FDA could not draw any scientific conclusions regarding selenium intake and the reduction of risk of any type of cancer from the animal or in vitro studies submitted with the petition.

C. Assessment of Intervention Studies

FDA evaluated 30 reports of intervention studies that were designed to evaluate the relationship between selenium intake and one or more of the cancers for which the petition requested a qualified health claim. Scientific conclusions could not be drawn from 29 of these 30 reports for one or more of the reasons discussed below (see also Appendix 1).[30]

Intervention Studies from Which No Scientific Conclusions Could Be Drawn

Bonelli et al. (1998) was an intervention study in which a selenium supplement that also contained zinc, vitamin A, vitamin C and vitamin E was given to Italian men and women with a history of adenomatous colorectal polyps. Colorectal carcinomas can grow from such polyps, although most polyps remain benign. This report stated that recruitment of study subjects was stopped at the end of 1995, and that the treatment period was 5 years. Thus, the results presented by Bonelli et al. (1998) are a preliminary report and it was stated in this report that the intervention was still ongoing for about one-fourth of the subjects.  FDA did not find any final, published reports of this study. Therefore, no scientific conclusions about the relationship between selenium intake and risk of colorectal cancer could be drawn from the study.

Two intervention studies were conducted on individuals who had prostate cancer (Hoenjet et al., 2005; Kranse et al., 2005). Health claims involve reducing the risk of a disease in people who do not have the disease that is the subject of the claim.  FDA considers evidence from studies with subjects who have the disease that is the subject of the claim only if it is scientifically appropriate to extrapolate to individuals who do not have the disease.  That is, the available scientific evidence must demonstrate that (1) the mechanism(s) for the mitigation or treatment effects measured in the diseased populations are the same as the mechanism(s) for risk reduction effects in non-diseased populations and (2) the substance affects these mechanisms in the same way in both diseased and healthy people. Because such evidence is not available for selenium and prostate cancer, the agency could not draw any scientific conclusions from studies that used subjects with prostate cancer to evaluate whether selenium reduces the risk of prostate cancer. 

Six intervention studies measured endpoints other than surrogate endpoints of a specific cancer risk. In general, surrogate endpoints are laboratory measurements of biomarkers (biological molecules found in blood, other body fluids, or tissues that are a sign of a normal or abnormal process, or of a condition or disease) that have been shown to be valid predictors of disease risk and therefore may be used in place of clinical measurements of the onset of the disease in a clinical trial (Spilker, 1991). Biomarkers that are associated with risk of a disease are called risk biomarkers.  All risk biomarkers for a given disease are also risk factors for that disease; however, not all risk factors for a disease are biomarkers. For instance, risk factors such as age and family history are not biomarkers.  Some risk biomarkers can be used to predict risk of a disease, while others are merely associated with risk of the disease and are not in the causal pathway leading to the disease (Biomarkers Definitions Working Group, 2001).  While there can be a number of risk biomarkers for a particular disease, very few have been validated for use as endpoints of disease risk.  In evaluating whether a biomarker or other study endpoint is a surrogate endpoint for risk of cancer, FDA ordinarily relies on the views of the National Institutes of Health's National Cancer Institute. 

The intervention trial reported by Prasad et al. (1995) and Krishnaswamy et al. (1995) was designed to examine the effects of a multiple-nutrient supplement (vitamin A, riboflavin, zinc and selenium) on potential biomarkers of genetic damage in cells scraped from inside the cheeks of Indian reverse smokers of chutta[31]. This study did not measure the incidence of oral cancer, but rather the frequency of micronucleated cells and carcinogen DNA adducts as indicators of DNA damage.  These are not surrogate endpoints of oral cancer risk.[32]  Three studies measured prostate-specific antigen (PSA) level in response to a supplement intervention (Hoenjet et al., 2005; Joniau et al., 2007; Kranse et al., 2005). While PSA tests are used as a first step in diagnosing prostate cancer, PSA level is not a surrogate endpoint for prostate cancer risk because elevated PSA levels are a result of various conditions (benign and malignant) of the prostate.[33]  Limburg et al. (2005) measured squamous dysplasia of the esophagus[34] in response to a supplement intervention. Squamous dysplasia of the esophagus occurs as a result of acid reflux, which may be associated with esophageal cancer in some people;[35] however, squamous dysplasia is not considered to be a validated surrogate endpoint for esophageal cancer.[36] Because these studies did not measure a validated surrogate endpoint, no scientific conclusions about the relationship between selenium intake and risk of prostate, oral, or esophageal cancer could be drawn from them.[37]

There were three reports of the SU.VI.MAX trial that evaluated the relationship between daily intake of supplements containing antioxidants and minerals (100 mg selenium, 120 mg vitamin C, 30 mg vitamin E, 6 mg beta-carotene, and 20 mg zinc) and risk of all cancers combined, as well as risk of the following site-specific cancers:  breast cancer and thyroid cancer (Hercberg et al., 2004), skin cancer (Hercberg et al., 2007), and prostate cancer (Meyer et al., 2005).  Because the petition did not request a qualified health claim for skin cancer, the agency did not review the report by Hercberg et al. (2007).  In the breast and thyroid cancer report by Hercberg et al. (2004), there were no diagnostic tests for the cancers of interest prior to the intervention. In the prostate cancer report, Meyer et al.  (2005) measured PSA levels at baseline; however, further tests for individuals with high PSA levels (≥ 4.0 µg/L) were left to the discretion of the individual's physician. Because the SU.VI.MAX study did not confirm that all subjects were free of the cancers of interest prior to the intervention, the study may have involved subjects who had the site-specific cancers evaluated in the two SU.VI.MAX reports, and consequently the results with respect to effects on the risk of those cancers may be biased.  Therefore, scientific conclusions could not be drawn from reports of this study about the relationship between selenium intake and cancer risk.

Eight reports discussed three intervention trials conducted in the Linxian region of China that examined the potential benefit of selenium in reducing the risk of esophageal, stomach, and liver cancer (Blot et al., 1993; Blot et al., 1995; Dawsey et al., 1994; Li et al., 1993; Limburg et al., 2005; Taylor et al. 1994; Qu et al., 2007; Wang et al., 1994). In the first trial, risk of esophageal, stomach, and liver cancer was evaluated in adults from the general population who received a daily placebo or a supplement containing α-tocopherol (vitamin E), β-carotene and selenium (Blot et al., 1993; Blot et al., 1995; Qu et al., 2007; Wang et al., 1994). In the second trial, risk of esophageal and stomach cancer was evaluated in adults from the general population who received a daily  placebo or supplement containing 26 vitamins and minerals (Blot et al., 1995; Dawsey et al., 1994;  Li et al., 1993; Taylor et al., 1994).  A third trial provided a daily placebo or 200 µg/day selenomethionine to subjects with mild or moderate esophageal squamous dysplasia (Limburg et al., 2005). Linxian, China has one of the world's highest rates of esophageal and gastric (stomach) cancer with mortality rates of these cancers exceeding the national U.S. average incidence by as much as 100-fold (Blot et al., 1993).  The reasons for the clustering of esophageal and stomach cancers in Linxian are unknown (Blot et al., 1993).  The prevalence of risk factors for stomach and esophageal cancer are significantly different between the United States and China.[38] There is a scarcity of food in the region, and this has led to sub-clinical deficiencies in several micronutrients, including vitamin A, in a substantial portion of the population (Yang et al., 1984; Yang et al. 1985).  Thus, in studies that provide a multinutrient supplement to residents of Linxian, it is not possible to determine whether a reduction in risk of any of the cancers evaluated in those who received the selenium-containing supplement is a result of correcting nutrient deficiencies or reflects a physiological effect that can be extrapolated to populations that do not have the same deficiencies. Thus, scientific conclusions about the role of selenium in reducing the risk of stomach, esophageal and liver cancer could not be drawn from this study because the study population had subclinical malnutrition and/or a strikingly high prevalence of esophageal and gastric cancer, as well as higher prevalence of risk factors than in the United States. [39]  

Two reports by Yu et al. (1991, 1997) on selenium supplements and primary liver cancer describe two separate trials conducted in Qidong County of China, which has an exceptionally high rate of this cancer. The reports note that the recognized risk factors potentially responsible for the high liver cancer rate in this county are aflatoxin contamination, hepatitis B viral infection, and water pollution.  The residents of Qidong County are nonmobile (i.e., most stay in the same geographic location throughout their lives) and consume mainly foods grown on local farms (Yu et al., 1985). The baseline blood selenium levels of the subjects in these trials were about 10 mg/dL, which is less than the 5th percentile of serum selenium levels in the United States (IOM, 2000). Because residents of Qidong consume much less selenium than the vast majority of the U.S. population and the prevalence of risk factors for liver cancer there is very different than in the United States, scientific conclusions could not be drawn from this study about the relationship between selenium intake and liver cancer in the general U.S. population.[40]

You et al. (2006) evaluated the effects of H. pylori treatment and long-term supplementation with a combination of vitamin C, vitamin E, and selenium in reducing the prevalence of precancerous gastric lesions, which are caused by H. pylori.  Li et al. (2004a) provided selenium to study subjects to evaluate the effects on risk of gastric cancer. These studies were conducted in Linqu or Qixia County, Shadong Province, China where the prevalence of precancerous gastric lesions is very high and the majority of adults have antibodies to H. pylori, indicating exposure to the bacteria.   Cancer is caused by external (e.g., diet, infections)[41] and internal factors[42] (e.g., genetics, hormones, immune function).  An estimated 50-80% of human cancers are caused by external factors.[43]  Different external or internal causal factors may alter the etiology of cancer in different populations.  The precise etiology of stomach cancer is unknown; however, one factor, H. pylori infection, is an external risk factor for stomach cancer, as it is associated with an increased risk of the disease.[44] The incidence of H. pylori infection is markedly higher in China than in the United States.[45] The existence of H. pylori infection creates bias,[46] because it can affect the risk of developing stomach cancer independent of selenium consumption.  Therefore, because of this external factor that may confound the results, study subjects from this region of China are not an appropriate population for determining whether selenium may reduce the risk of stomach cancer in the U.S. population.  Accordingly, results of studies on this Chinese population cannot be extrapolated to reach conclusions about the effects of selenium on the U.S. population.

One intervention study measured colorectal polyp recurrence in Norwegians given a placebo or a supplement containing selenium (101 µg/day), β-carotene, vitamin E, vitamin C and calcium daily for 3 years (Hofstad et al., 1998). Limited credible evidence has been reported for both calcium[47] and vitamin E[48] and reduced risk of colorectal cancer in the general U.S. population.  Because the test supplement contained calcium and vitamin E as well as selenium, it is not possible to accurately determine whether any observed effects on risk of colorectal cancer in the study were due to selenium, or to calcium and vitamin E acting alone or together.[49] Therefore, this study cannot be used to evaluate the independent effect of selenium on risk of colorectal cancer. 

There were seven reports of one trial, the Nutritional Prevention of Cancer (NPC) Trial, that evaluated the relationship between supplemental selenium and skin cancer risk (Clark et al., 1996; Clark et al., 1998; Duffield-Lillico et al., 2002; Duffield-Lillico et al., 2003a, 2003b; Reid et al., 2002; Reid et al., 2006). The NPC trial was a double-blind, randomized trial that supplemented men and women (n=1,312) on the east coast of the United States with a placebo or 200 µg selenium/day (supplied as selenium-enriched yeast, which approximates the forms of selenium found in the diet).  Subjects were eligible to participate in the study if they had a history of two or more basal cell carcinomas or one squamous cell carcinoma within the prior year.  The NPC trial continued for one year beyond the interim reports, and findings for the entire duration of the trial are published in four reports (Duffield-Lillico et al., 2002; Duffield-Lillico et al., 2003a, b; Reid et al., 2002).   While non-melanoma skin cancer was the primary outcome measured (Duffield-Lillico et al., 2003b), post-hoc analyses were conducted to examine other (secondary) cancers (prostate, lung, breast, colorectal, ovarian, bladder, head and neck, and lymphatic),  which the study was not designed to evaluate (Clark et al., 1996; Clark et al., 1998; Duffield-Lillico et al., 2002; Duffield-Lillico et al., 2003a; Reid et al., 2002; Reid et al., 2006). Consequently, the results with respect to cancers other than skin cancer[50] may be biased due to preexisting cases of the cancer of interest in the treatment or placebo group.  Therefore, scientific conclusions about selenium supplements and secondary cancers could not be drawn from the seven reports since the study did not confirm that all subjects were free of the cancers of interest prior to the intervention.

Based on the above discussion, there was one intervention study available from which scientific conclusions could be drawn about the relationship between selenium intake and risk of a site-specific cancer for which the petition requested a qualified health claim.  This was a study by Lippman et al. (2009) that evaluated the relationship between selenium supplements and prostate cancer risk.   

Bladder and Urinary Tract Cancers

There were no intervention studies from which scientific conclusions could be drawn about the relationship between selenium supplements and risk of bladder cancer and no intervention studies about the relationship between selenium supplements and risk of other cancers of the urinary tract.

Brain Cancer

There were no intervention studies about the relationship between selenium supplements and risk of brain cancer.

Breast Cancer

There were no intervention studies from which scientific conclusions could be drawn about the relationship between selenium supplements and risk of breast cancer.

Colon and Digestive Tract Cancers

There were no intervention studies from which scientific conclusions could be drawn about the relationship between selenium supplements and risk of colon cancer or other cancers of the digestive tract.

Liver Cancer

There were no intervention studies from which scientific conclusions could be drawn about the relationship between selenium supplements and risk of liver cancer.

Lung and Respiratory Tract Cancers

There were no intervention studies from which scientific conclusions could be drawn about the relationship between selenium supplements and risk of lung cancer or other cancers of the respiratory tract.

Prostate Cancer

 Lippman et al. (2009) reported on the Selenium and Vitamin E Cancer Prevention Trial, an intervention study of moderate methodological quality in which the subjects were men over 50 years of age. In the selenium-only arm of this study, the treatment group (n=8,752) received a selenium supplement (200 µg/day of selenomethionine), while the control group (n=8,696) received a placebo.  Prostate cancer was the primary endpoint, with results of PSA and digital rectal examination required at the study entry to screen out men who already had prostate cancer.[51]  Serum selenium levels were higher for the selenium group (198-275 µg/L) than for the placebo group (120-152 µg/L), which indicated that the selenium treatment group was taking the provided supplements. During the 5 year follow-up period, there were 416 cases of prostate cancer in the placebo group and 432 cases in the selenium group. There was no significant difference in the 5 year incidence of prostate cancer between the placebo (4.43%) and the selenium (4.56%) group.  Lippman et al. (2009) also evaluated lung and colorectal cancer as secondary cancers. However, for the reasons outlined in the discussion of the Nutritional Prevention of Cancer Trial above, no scientific conclusions about selenium supplements and lung or colorectal cancer could be drawn from this study since it did not screen for lung and colorectal cancer prior to the intervention.

Thyroid Cancer

There were no intervention studies from which scientific conclusions could be drawn about the relationship between selenium supplements and risk of thyroid cancer.

D. Assessment of Observational Studies

Assessment of Selenium Intake

Dietary Selenium

The majority (> 80%) of selenium in conventional foods (e.g., corn, wheat and soybeans) is present as L-selenomethionine (Yang et al., 1997). The bioavailability of selenomethionine is greater than 90 percent (Thomson and Robinson, 1986). Selenocysteine, another organic form of dietary selenium, is also highly bioavailable (Swanson et al., 1991). The bioavailability of inorganic forms of selenium (e.g., selenate and selenite) is similar to selenomethionine (Dresoti, 1986), but there is evidence to suggest that selenate absorption is reduced in the presence of food (50 to 80 percent) (Thomson and Robinson, 1986).  Overall, absorption of selenium is relatively high from conventional foods and from chemically pure forms, and varies with the form of selenium and selenium status of the subject. (Finley et al., 2006).  In 1984, synthetic selenomethionine became available for use in dietary supplements. Currently, selenium-containing dietary supplements are sold most often as sodium selenate or selenomethionine.

In observational studies that calculate nutrient intake from conventional foods, measures of selenium intake are based on recorded dietary intake methods such as food frequency questionnaires, diet recalls, or food records, in which the type and amount of foods consumed are estimated.  A common weakness of observational studies is the limited ability to ascertain the actual food or nutrient intake for the population studied as a result of poor memory, over- or underestimation of portion sizes, and recall bias (Flegal, 1999).  Furthermore, the nutrient content of foods can vary (e.g., due to soil composition, food processing and cooking procedures, or storage conditions (duration, temperature)).  The amount of selenium in a specific food is dependent on the availability of selenium from the soil where the raw agricultural product is produced (Combs, 2000), and similar food products may contain concentrations of selenium that vary by 10-fold or more (Finley et al., 1996).  Dietary questionnaires are not capable of measuring selenium intake because of the highly variable composition of foods (Hunter et al., 1990b). Thus, to accurately estimate intake of selenium requires direct chemical analysis of the diet consumed, which is not done when assessing selenium intake in observational studies. Because of the high variability of selenium in specific foods, methods for assessing dietary intake are not considered to be reliable for assessing selenium intake (Clark, 1985; Levander, 1985). Thus, it is difficult to ascertain an accurate amount of selenium consumed based on reports of dietary intake of foods. For these reasons, most observational studies measure a biomarker of selenium intake rather than estimating intake by dietary assessment methods.

In addition, conventional foods contain not only selenium, but also other nutrients that may be associated with the metabolism of selenium or the pathogenesis of various cancers.  Because foods consist of many nutrients and substances, it is difficult to study a nutrient or food component in isolation (Sempos et al., 1999).  (See Sempos et al. (1999), Willett (1990), and Willett (1998) regarding the complexity of identifying the relationship between a specific nutrient within a food and a disease). For studies based on recorded dietary intake of such foods, it is not possible to accurately determine whether any observed effects of selenium on cancer risk were due to: 1) selenium alone; 2) interactions between selenium and other nutrients; 3) other nutrients acting alone or together; or 4) decreased consumption of other nutrients or substances contained in foods displaced from the diet by the increased intake of selenium-rich foods.

In fact, evidence demonstrates that in a number of instances, epidemiological studies based on the recorded dietary intake of conventional foods may indicate a benefit for a particular nutrient with respect to a disease, but it is subsequently demonstrated in an intervention study that the nutrient-containing dietary supplement does not confer a benefit or actually increases risk of the disease (Lichtenstein and Russell, 2005).  For example, previous epidemiological studies reported an association between fruits and vegetables high in beta-carotene and a reduced risk of lung cancer (Peto et al., 1981).  However, subsequent intervention studies, the Alpha-Tocopherol and Beta Carotene Prevention Study (ATBC) and the Carotene and Retinol Efficiency Trial (CARET), demonstrated that beta-carotene supplements increase the risk of lung cancer in smokers and asbestos-exposed workers, respectively (The Alpha-Tocopherol and Beta Carotene Cancer Prevention Study Group, 1994; Omenn et al., 1996).  These studies illustrate that a nutrient provided as a dietary supplement may exhibit different health effects compared to when it is consumed among many other food components.  Furthermore, these studies demonstrate the potential public health risk of relying on results from epidemiological studies, in which the effect of a nutrient is based on recorded dietary intake of conventional foods as the sole source for concluding that a relationship exists between a specific nutrient and disease risk; the effect could actually be harmful. 

In Pearson v. Shalala, the D.C. Circuit noted that FDA had "logically determined" that the consumption of a dietary supplement containing antioxidants could not be scientifically proven to reduce the risk of cancer where the existing research had examined only foods containing antioxidants, as the effect of those foods on reducing the risk of cancer may have resulted from other substances in those foods (164 F.3d 650, 658 (D.C. Cir 1999)).  The D.C. Circuit, however, concluded that FDA's concern with granting antioxidant vitamins a qualified health claim could be accommodated by simply adding a prominent disclaimer noting that the evidence for such a claim was inconclusive, given that the studies supporting the claim were based on foods containing other substances that might actually be responsible for reducing the risk of cancer.  Id.  The court noted that FDA did not assert that the dietary supplements at issue would "threaten consumer's health and safety."  Id. at 656.  There is, however, a more fundamental problem with allowing qualified health claims for nutrients in dietary supplements based solely on studies of foods containing those nutrients than the problem the D.C. Circuit held could be cured with a disclaimer.  As noted above,  even if the effect of the specific component of the food constituting the dietary supplement could be determined with certainty, recent scientific studies have shown that nutrients in food do not necessarily have the same beneficial effect when taken in the form of a dietary supplement.  See Lichtenstein and Russell (2005).  Indeed, not only have studies on single nutrient supplements established that the benefits associated with the dietary intake of certain nutrients do not materialize when the nutrients are taken as a supplement, but some of these studies have actually indicated an increased risk for the very disease the nutrients were predicted to prevent.  Id.  Thus, an observational study based on intake of conventional food provides no information from which scientific conclusions may be drawn for the nutrient when taken as a supplement. 

Therefore, observational studies based on intake of conventional foods do not provide any credible evidence for a risk reduction claim about a nutrient taken as a supplement because, in fact, the nutrient in supplement form may decrease, have no effect, or actually increase risk of the disease or health-related condition.  For the reasons set forth in Section V., entitled "Agency's Consideration of Disclaimers or Qualifying Language," FDA has concluded that neither a disclaimer nor qualifying language would suffice to prevent consumer deception in these instances because observational studies in food do not provide credible evidence of risk reduction for a nutrient provided in a dietary supplement.

Blood selenium concentration

Plasma and serum[52] selenium have been shown to correlate with selenium intake (r[53] = 0.63-0.97) (Longnecker et al., 1991; Longnecker et al., 1996; Swanson et al., 1990; Yang et al., 1989). Selenium intake has been shown to be the strongest predictor of serum selenium (Swanson et al., 1990).  Biomarkers that are reliable estimates of intake of a nutrient, such as selenium, are not affected by the limitations of estimating food intake in observational studies as discussed above.  Nutritional adequacy of selenium is defined as the amount of selenium that is needed to saturate selenium enzyme activities; this occurs at blood selenium concentrations of 70-90 µg/L (Thomson, 2004). Using data from NHANES III, the distribution of average serum selenium concentrations in the U.S. population was reported to range from 91 mg/L (1st percentile) to 168 mg/L (99th percentile) (Niskar et al., 2003).  Consequently, populations with blood selenium concentrations of less than 90 µg/L are very different from the population in the United States, and a substantial proportion of individuals in these populations are likely to be selenium-deficient.  Therefore, studies of populations where the majority of the subjects have blood selenium concentrations of less than 90 µg/L are not applicable to the general U.S. population.  

Two chemical forms account for nearly all of the selenium in the plasma and serum components of the blood: selenomethionine and selenocysteine. The bioavailability of organic and inorganic selenium from conventional foods and chemically pure forms of the mineral (e.g., those used in dietary supplements) is high. However, there are a number of studies to suggest that the metabolism of selenomethionine and other forms of selenium is different, such that the other forms do not raise plasma and serum selenium levels to the same degree as selenomethionine, or not at all (Burk et al., 2001; Burk et al., 2006; Butler et al., 1991; Moser-Veillon et al., 1992; Robinson et al., 1997; Thomson et al., 1993; Veillon et al., 1990; Burk and Levander, 2006). This is because selenomethionine binds to various proteins in the blood (selenoproteins), whereas selenocysteine and inorganic selenium, once absorbed, are transported to the liver, where these compounds enter the selenium metabolic pool and are distributed throughout the body ( Burk and Levander, 2006).  Therefore, studies that use plasma or serum selenium as a biomarker of intake are considered to reflect the intake of selenomethionine.

Toenail selenium concentration

Ecological studies have demonstrated the utility of toenail selenium as a useful biomarker of selenium intake for comparing populations of differing selenium intake (Morris et al., 1983; Xia et al., 2000). Toenail selenium levels are unaffected by recent selenium intake (less than 9 months of measurement (Longnecker et al., 1993)), and are a useful biomarker of long-term selenium intake. The correlation coefficient (r) between dietary selenium[54] (based on duplicate-plate food collections)[55]and toenail selenium for 142 subjects in areas of South Dakota and eastern Wyoming where the soil is rich in selenium was moderately high (r =0.69) (Longnecker et al., 1991). The correlation between toenail selenium and serum selenium concentration was high (r=0.89). A subsample (n=77) of the Longnecker et al. (1991) study showed that toenail selenium levels correlated moderately well (r=0.67) with selenium intake assessed by duplicate-plate collections; adjusting for differences in calorie intake and non-dietary covariates (e.g., smoking, age, and gender) improved the ability of toenail selenium concentration to predict selenium intake (r = 0.72) (Longnecker et al., 1996). It has been demonstrated that smoking significantly lowers toenail selenium levels, that smoking is an independent determinant of toenail selenium concentration (Swanson et al., 1990; Krogh et al., 2003), and that it is necessary to adjust for smoking in observational studies (Hunter et al., 1990b).   Based on the above, toenail selenium is considered to be a reliable biomarker of selenium intake in observational studies, provided that the data are adjusted for smoking. Because organic and inorganic selenium are well absorbed from conventional foods and chemically pure forms of selenium (e.g., those used in dietary supplements), toenail selenium can be used as a biomarker of total selenium intake from both conventional foods and dietary supplements.

As noted above, the form of selenium consumed affects plasma and serum selenium levels.  The available evidence suggests that the same is not true for toenail selenium levels, however.  As discussed above, selenomethionine is the only form of selenium that binds to protein in blood.  Other forms of selenium, including selenocysteine and inorganic forms such as selenate and selenite, are not converted to selenomethionine upon absorption (Burk and Levander, 2006). Therefore, these other ingested forms of selenium are cleared faster from blood than selenomethionine.  Once cleared from the blood, all forms of selenium are distributed throughout the body and deposited in tissue, including the nails (Burk and Levander, 2006). 

The U.S. Health Professionals follow-up study reported an average toenail selenium level of 0.87 mg/g (n= 935) in men, with the lowest quintile of toenail selenium level ranging from 0.66 to 1.14 mg/g (Yoshizawa et al., 1998).  Garland et al. (1995) reported that the average toenail selenium concentration from 503 middle-aged U.S. female nurses was 0.86 mg/g, with the lowest quintile of toenail selenium level representing less than 0.72 mg/g.

Observational Studies

Based on the above discussions, prospective observational studies that measured plasma, serum, or toenail selenium as biomarkers of selenium intake were used for making scientific conclusions about the relationship between selenium intake and risk of site-specific cancers. Although observational studies that use dietary assessment methods to calculate intake of selenium from conventional foods provide no information from which scientific conclusions may be drawn for the nutrient when taken as a supplement, these biomarkers of selenium intake can reflect intake of selenium from either conventional foods or dietary supplements.

FDA reviewed a total of 105 observational studies that evaluated the relationship between selenium and cancer risk. Scientific conclusions could not be drawn from 87 of the observational studies for one or more of the reasons discussed below (see also Appendix 1).

Seventeen observational studies were conducted in selenium-deficient populations (Appendix 1). Two were prospective cohort studies conducted in Linxian, China.  These studies evaluated the relationship between serum selenium levels and risk of esophageal and stomach cancer (Mark et al., 2000; Wei et al., 2004).  In one study conducted in Linxian, the average serum selenium concentration was 72.2 ± 0.57 mg/L, and 89% of the subjects had a serum selenium level less than 92.8 mg/L at the start of the study (Mark et al., 2000).  Fifteen studies were conducted in Finland or the Netherlands, where baseline selenium levels were low at the time the studies were conducted (Hartman et al., 2002; Knekt et al. 1988; Knekt et al., 1991; Knekt et al., 1990, Knekt et al., 1998; Michaud et al., 2002; Salonen et al., 1984; Salonen et al., 1985; Van't Veer et al., 1996; Van den Brandt et al., 1993a; Van den Brandt et al., 1993b; Van den Brandt et al., 1994; Van den Brandt et al., 2003; Virtamo et al., 1987; Zeegers et al., 2002). For example, the average baseline serum selenium levels in one Finnish study were 48.6 mg/L for gastrointestinal cancer cases and 54.3 mg/L for controls (Salonen et al., 1984).  Similarly, the average baseline serum selenium levels in two other Finnish studies were only 53.9 to 62.5 µg/L for both cases and controls (Knekt et al., 1990; Virtamo et al., 1987). For Dutch subjects, baseline toenail selenium levels were approximately 0.5 µg/g (van den Brandt et al., 1993; van den Brandt et al., 1994). As discussed above under the headings "Blood selenium concentration" and "Toenail selenium concentration," the serum and toenail selenium levels in the vast majority of the U.S. population are significantly higher than observed in these Chinese, Finnish, or Dutch subjects. It cannot be determined whether a relationship between selenium intake and cancer risk in the Chinese, Dutch, and Finnish populations would be observed if the baseline serum or toenail selenium levels were similar to those in the U.S. population. Thus, scientific conclusions cannot be drawn from these studies about a relationship between selenium intake and cancer risk in the general U.S. population.

Two studies, one conducted in China and the other in Taiwan, evaluated the relationship between selenium intake and liver cancer in populations where the prevalence of hepatitis B and liver cancer is extremely high (Sakoda et al., 2005; Yu et al., 1999). Because hepatitis B is a risk factor for liver cancer,[56] it is not possible to draw conclusions from these studies about the relationship between selenium intake and liver cancer in the general U.S. population, where the prevalence of hepatitis B is low.[57]

One study was a nested case-control[58] study that followed Japanese survivors of the atomic bomb in 1945 (Kabuto et al., 1994). This study evaluated the relationship between serum selenium levels and the risk of stomach and lung cancer. The incidence of stomach and lung cancer among atomic bomb survivors is related to radiation exposure (Yamamoto et al., 1986). The extreme level of radiation exposure experienced by the atomic bomb survivors in Japan has not been experienced by inhabitants of the United States. Therefore, it is not possible to determine how a relationship between selenium intake and risk of lung cancer in atomic bomb survivors could be extrapolated to the general U.S. population. Thus, scientific conclusions could not be drawn from this study about the relationship between selenium intake and risk of stomach or lung cancer in the U.S. population. 

Three observational studies evaluated the relationship between selenium intake and deaths from all cancers combined (Fex et al., 1987; Kok et al., 1987; Kornitzer et al., 2004). The studies did not evaluate the relationship between selenium intake and risk of individual cancers. As stated earlier, cancer is a constellation of more than 100 different diseases, each of which is characterized by the uncontrolled growth and spread of abnormal cells (American Cancer Society, 2008).  Cancers at different organ sites have different risk factors, treatment modalities, and mortality risk (American Cancer Society, 2008).  Both genetic and environmental risk factors may affect the risk of different types of cancers.  Risk factors may include a family history of a specific type of cancer, cigarette smoking, alcohol consumption, overweight and obesity, exposure to ultraviolet or ionizing radiation, exposure to cancer-causing chemicals, and dietary factors.  The etiology, risk factors, diagnosis, and treatment for each type of cancer are different (Hord and Fenton, 2007; Milner, 2006).  Since each form of cancer is different, FDA evaluates each cancer individually.  However, these studies looked at all cancers combined and did not report the results for the individual cancers included in the evaluation, making it impossible to know which cancers selenium had an effect on.  Further, these studies did not match or adjust for the various confounders of risk for each individual cancer.  Known confounders of disease risk need to be collected and adjusted for to minimize bias. For example, information on each subject's risk factors, such as age, race, body weight and smoking, should be collected and used to adjust the data so that the substance/disease relationship is accurately measured.[59]  Because these three studies did not evaluate the effect of selenium on risk of any site-specific cancers and did not adjust for confounders of risk, no scientific conclusions about the relationship between selenium intake and risk of any site-specific cancers could be drawn from these three studies.[60]

Eleven studies estimated selenium intake by using dietary assessment methods to estimate the types and quantities of foods consumed by the study subjects and then using food composition tables to calculate the amount of selenium in the foods consumed (Appendix 1). As discussed above, the concentration of selenium in a food depends on where the food was grown and the accuracy of food composition tables. Therefore, dietary assessment methods do not provide a reliable estimate of dietary selenium intake. Also, as previously discussed, benefits associated with the dietary intake of certain nutrients do not necessarily materialize when the nutrients are taken as supplements.  Thus, scientific conclusions about the relationship between selenium intake and specific cancers could not be drawn from these eleven studies.

One study evaluated the relationship between selenium intake from drinking water and various cancers (Vinceti et al., 2000). While the range of selenium concentration in drinking water was known for the region where the study took place, there was no information collected on the study subjects' consumption of drinking water. Therefore, the amount of selenium the subjects consumed from drinking water was not determined in this study.  In addition, drinking water may contain minerals and contaminants that are associated with the risk of cancer. Therefore, this study cannot be used to evaluate the independent association between selenium and risk of the cancers studied, nor can it be used to draw any scientific conclusions about the relationship between selenium intake and risk of these cancers.

Three studies did not conduct appropriate statistical analysis (Coates et al., 1988; Nomura et al., 1987; Peleg et al., 1985). These three nested case-control studies did not provide confidence intervals (CI)[61] for determining statistical significance of risk between different levels of serum selenium. Statistical analysis of the study findings is a critical factor because it provides the comparison in cancer risk between subjects consuming selenium and those not consuming selenium.   Thus, when appropriate statistical analysis is not performed, it is not possible to determine whether the results showed a difference in risk between the two study groups.  As a result, these studies provided no information about whether selenium may reduce the risk of cancer, and no scientific conclusions could be drawn from them.

Four studies did not adequately adjust for confounders of risk of the specific type of cancer being studied (Helzlsouer et al., 2000; Van Noord et al., 1987; Van Noord et al., 1993; Yoshizawa et al., 1998), and one study did not adequately adjust for smoking, which is a confounder of toenail selenium levels (Garland et al., 1995). Known confounders of disease risk and biomarkers of intake need to be collected and adjusted for to minimize bias. For example, information on each subject's risk factors, such as age, race, body weight and smoking, should be collected and used to adjust the data so that the substance/disease relationship is accurately measured.[62]

Thirty-six studies were retrospective case-control or cross-sectional studies that measured a biomarker of selenium intake (Appendix 1).  In retrospective studies, such measurements are taken after the cases (subjects who have the disease being studied) have already been diagnosed.  The health status of study subjects can have an effect on nutrient biomarkers, including selenium. Cancer induces widespread cellular and metabolic alterations that affect lipids, proteins, carbohydrates, vitamins, minerals and hormones (Schattner and Shike, 2006). Thus, it is not possible to determine whether altered selenium biomarker levels in a retrospective study are due to having cancer or due to nutrient intake in the above case-control and cross-sectional studies. For these reasons, scientific conclusions about the relationship between selenium intake and risk of cancer in the general U.S. population could not be drawn from these 36 studies.[63]

The following 19 observational studies 1) used reliable biomarkers of selenium intake (i.e., serum, plasma, or toenail selenium concentration), 2) were  nested case-control studies, such that plasma, serum or toenail selenium was measured prior to the diagnosis of the cancer being studied, and 3) were designed such that scientific conclusions could be drawn about the relationship between selenium intake and risk of site-specific cancers (e.g., matched or adjusted for confounders of disease risk).[64]

Bladder and Urinary Tract Cancers

In the group of studies from which scientific conclusions could be drawn, FDA identified two articles on a total of three prospective observational studies that evaluated the relationship between serum or toenail selenium concentration and risk of bladder cancer (Helzlsouer et al., 1989; Michaud et al., 2005). There were no observational studies on selenium intake and risk of urinary tract cancers other than bladder cancer.

Helzlsouer et al. (1989) was a nested case-control study of high methodological quality that followed 20,305 U.S. men and women for 12 years. Average serum selenium levels were significantly lower in 35 bladder cancer cases (111 µg/L) compared to 70 matched controls (117 µg/L) (p = 0.03)[65]. However, there was no significant difference in the observed risk for serum selenium levels and bladder cancer (odds ratio[66] = 0.48; CI = 0.67-6.35).  Therefore, this study does not provide any evidence for a relationship between selenium intake and reduced risk of bladder cancer.   

Michaud et al. (2005) reported on the results of two nested case-control studies. Both studies were of high methodological quality. One of the studies followed 33,737 men for 13 years (Health Professionals Follow-up Study) and the other study followed 68,213 women for 17 years (Nurses' Health Study).  Based on 222 bladder cancer cases and 224 controls, there was no significant relationship between toenail selenium levels and risk of bladder cancer in men (relative risk[67] = 1.17; CI = 0.66-2.07). There was, however, a negative relationship between toenail selenium levels and bladder cancer risk (i.e., the higher the level of toenail selenium, the lower the risk of bladder cancer) in women based on 116 bladder cancer cases and 117 controls (relative risk = 0.36; CI = 0.14-0.91).

Brain Cancer

There were no observational studies from which scientific conclusions could be drawn about the relationship between selenium intake and risk of brain cancer.

Breast Cancer

In the group of studies from which scientific conclusions could be drawn, FDA identified six prospective observational studies that reported on the relationship between plasma, serum, or toenail selenium and breast cancer (Criqui et al., 1991; Dorgan et al., 1998; Hunter et al., 1990a; Overvad et al., 1991; Ringstad et al., 1988; Willett et al., 1983). 

Criqui et al. (1991) was a nested case-control study of moderate methodological quality that followed 8,825 U.S. subjects for 8.5 years. There was no significant difference in the plasma selenium levels between breast cancer cases (n=16) and controls (n=30).  Therefore, this study does not provide any evidence for a relationship between selenium and reduced risk of breast cancer.

Dorgan et al. (1998) was a nested case-control study of moderate methodological quality that followed 7,224 U.S. women for up to 9.5 years. Based on 105 cases and 203 controls, there was no significant association between serum selenium levels and risk of breast cancer (relative risk = 0.9; CI=0.3-2.5).

Hunter et al. (1990a) was a nested case-control study of high methodological quality that followed 62,641 U.S. women for 53 months. Based on 434 breast cancer cases and 434 controls, there was no association between toenail selenium levels and risk of breast cancer (relative risk = 1.1; CI=0.70-1.72).

Overvad et al. (1991) was a nested case-control study of moderate methodological quality. After following 5,162 women on the Island of Guernsey for an average of 11 years, 46 breast cancer cases were diagnosed. There was no significant difference in plasma selenium levels between the breast cancer cases (109 µg/L) and the matched controls (103 µg/L). Furthermore, there was no significant relationship between plasma selenium levels and risk of breast cancer (odds ratio = 0.80; CI =0.29-2.19)

Ringstad et al. (1988) was a nested case-control study of moderate methodological quality that followed 9,364 Norwegian women for 6 years. There was no significant difference in serum selenium levels between 10 breast cancer cases (124 mg/L) and 10 controls (135 mg/L).  Therefore, this study does not provide any evidence for a relationship between selenium and reduced risk of breast cancer.

Willett et al. (1983) was a nested case-control study of moderate methodological quality that followed 10,940 U.S. women for 5 years. There was no significant difference between the serum selenium levels of the 16 controls (136 µg/L) and the 16 breast cancer cases (135 µg/L). Therefore, this study does not provide any evidence for a relationship between selenium and reduced risk of breast cancer.

Colon and Digestive Tract Cancers

In the group of studies from which scientific conclusions could be drawn, the agency identified three prospective observational studies that reported on serum selenium concentration and the risk of colorectal cancer (Peters et al., 2006; Schober et al. 1987; Wallace et al., 2003). There were no studies from which scientific conclusions could be drawn about the relationship between selenium intake and risk of digestive tract cancers other than colorectal cancer.

Three studies discussed in the previous section on breast cancer studies also evaluated the relationship between plasma or serum selenium and gastrointestinal cancer (Criqui et al. 1991; Ringstad et al., 1988; Willett et al., 1983). However, gastrointestinal cancer includes a number of site-specific cancers that have different etiologies and risk factors.[68] These three studies did not specify the individual site-specific cancers included as gastrointestinal cancers or report results for those individual cancers.  Therefore, no scientific conclusions about the relationship between selenium intake and risk of individual digestive tract cancers could be drawn from these three studies.

Peters et al. (2006) was a U.S. nested case-control study from the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial. This study was of high methodological quality and included 758 cases of colorectal polyps and 767 sex- and race-matched controls. There was no significant difference in the overall incidence of adenomatous colorectal polyps (a surrogate endpoint for colorectal cancer) between the five quintiles of serum selenium. Therefore, this study does not provide any evidence for a relationship between selenium and reduced risk of colorectal cancer.

Schober et al. (1987) was a U.S. nested case-control study of high methodological quality that measured pre-diagnostic (i.e., before diagnosis of cancer) serum selenium levels in 72 colorectal cancer cases and 143 controls. There was no significant difference in serum selenium levels between the cases (110 µg/L) and the controls (115 µg/L). Furthermore, there was no significant relationship between serum selenium levels and risk of colorectal cancer (odds ratio = 1.1; CI = 0.5-2.4).

Wallace et al. (2003) was a U.S. nested case-control study of high methodological quality that measured pre-diagnostic serum selenium levels in 276 cases (defined as individuals who later developed a colorectal cancer) and 276 controls. There was no significant relationship between serum selenium levels and risk of recurrent colorectal polyps (odds ratio =0.76; CI = 0.44-1.3).

Liver Cancer

There were no observational studies from which scientific conclusions could be drawn about the relationship between selenium intake and risk of liver cancer.

Lung and Respiratory Tract Cancers

In the group of studies from which scientific conclusions could be drawn, FDA identified four prospective observational studies that reported on the relationship between serum selenium and risk of lung cancer (Goodman et al., 2001; Menkes et al., 1986; Ringstad et al., 1988; Willett et al., 1983). There were no observational studies about the relationship between selenium intake and risk of respiratory tract cancers other than lung cancer.

Goodman et al. (2001) was a nested case-control study of high methodological quality that followed 14,254 U.S. smokers for 4.7 years. There was no significant difference in mean serum selenium concentrations between 456 controls (118 µg/L) and 356 lung cancer cases (119 µg/L). Furthermore, there was no significant relationship between serum selenium concentration and lung cancer risk (p = 0.49).

Menkes et al. (1986) was a nested case-control study of high methodological quality that followed 25,802 U.S. men and women for up to 9 years. There was no significant difference in mean serum selenium levels between the 99 lung cancer cases (113 µg/L) and 196 controls (110 µg/L). Therefore, this study does not provide any evidence for a relationship between selenium and reduced risk of lung cancer.

Ringstad et al. (1988) was a nested case-control study of high methodological quality that followed 9,364 Norwegian men and women for 6 years. There was no significant difference in mean serum selenium levels between the seven lung cancer cases (124 mg/L) and seven controls (128 mg/L). Therefore, this study does not provide any evidence for a relationship between selenium and reduced risk of lung cancer.

Willett et al. (1983) was a nested case-control study of high methodological quality that followed 10,940 U.S. hypertensive men and women for 5 years. There was no significant difference in the mean serum selenium levels between the 18 controls (131 µg/L) and 18 lung cancer cases (122 µg/L). Therefore, this study does not provide any evidence for a relationship between selenium and reduced risk of lung cancer.

Prostate Cancer 

In the group of studies from which scientific conclusions could be drawn, FDA identified eight prospective observational studies that reported on the relationship between plasma or serum selenium and risk of prostate cancer (Brooks et al., 2001; Criqui et al., 1991; Goodman et al., 2001; Li et al., 2004b; Nomura et al., 2000; Peters et al., 2007; Peters et al., 2008; Willett et al., 1983). 

Brooks et al. (2001) was a nested case-control study of moderate methodological quality that followed 1,555 U.S. men for ten years and compared plasma selenium concentrations in 52 men diagnosed with prostate cancer to 96 matched cancer-free controls.  There was no significant difference in plasma selenium concentrations between the prostate cancer cases (122 µg/L) and controls (117 µg/L). However, there was a significant negative relationship between selenium status and risk of prostate cancer (odds ratio = 0.24; CI=0.07-0.77), suggesting that selenium may reduce the risk of prostate cancer.      

Criqui et al. (1991) was a nested case-control study of moderate methodological quality that followed 8,825 men and women for 8.5 years. The total number of men followed for the purpose of identifying prostate cancer cases was not reported. There was no significant difference in plasma selenium levels between six prostate cancer cases and 12 controls. Therefore, this study does not provide any evidence for a relationship between selenium and reduced risk of prostate cancer.

Goodman et al. (2001) was a nested case-control study of moderate methodological quality that followed 14,254 U.S. smokers for 4.7 years. There was no significant difference in the serum selenium concentrations between 456 cancer-free controls (118 µg/L) and 235 prostate cancer cases (119 µg/L). Furthermore, there was no significant difference in the odds ratio for prostate cancer risk between the four quartiles of serum selenium levels (odds ratio = 1.2; CI =0.77-1.88).

Li et al. (2004b) was a nested case-control study of moderate methodological quality that followed 22,071 U.S. men for 13 years.  Plasma selenium levels from 586 prostate cancer cases were compared to 577 matched cancer-free cohort controls.  There was no significant difference in plasma selenium levels between the cancer-free controls (108 µg/L) and prostate cancer cases (106 µg/L).  Furthermore, there was no significant relationship between plasma selenium levels and prostate cancer risk (odds ratio = 0.78; CI = 0.54-1.13).

Nomura et al. (2000) was a nested case-control study of moderate methodological quality that followed 9,345 Japanese-American men living in Hawaii for 23 years.  The serum selenium concentration of 249 prostate cancer cases (130 µg/L) was significantly lower than that of the 249 matched cancer-free controls (134 µg/L). Furthermore, there was a significant negative relationship between serum selenium and risk of prostate cancer (odds ratio = 0.5; CI = 0.3-0.9), suggesting that selenium may reduce the risk of prostate cancer.

Peters et al. (2007) was a nested case-control study of high methodological quality that followed 26,975 men in 10 different U.S. cities for 8 years. Based on 724 prostate cancer cases and 879 controls, there was no significant difference in the odds ratio (0.84; CI= 0.62-1.14) for prostate cancer risk and the four quartiles of serum selenium concentration.

Peters et al. (2008) was a nested case-control study of high methodological quality that followed 35,242 U.S. men for 10 years. Based on 818 cases and 34,600 controls, no association was observed between current or past intake of single-ingredient selenium supplements and prostate cancer risk (hazard ratio[69] 0.90; CI= 0.62-1.3) for a 10-year average supplemental intake greater than 50 µg/day versus non-use of selenium supplements.

Willett et al. (1983) was a nested case-control study of moderate methodological quality that followed 4,480 U.S. hypertensive men for 5 years. There was no significant difference in the average serum selenium level between the 11 controls (139 µg/L) and 11 prostate cancer cases (128 µg/L).  Therefore, this study does not provide any evidence for a relationship between selenium and reduced risk of prostate cancer.

Thyroid Cancer

In the group of studies from which scientific conclusions could be drawn, FDA identified one prospective observational study that reported on the relationship between serum selenium and thyroid cancer (Glattre et al., 1989).

Glattre et al. (1989) was a nested case-control study of moderate methodological quality that followed approximately 100,000 Norwegian men and women from four different counties for 14 years. The mean serum selenium concentration of 43 thyroid cancer patients (108 mg/L) was significantly lower than the mean serum selenium concentration in 129 matched controls (115 mg/L) ( p < 0.025).  When serum selenium levels were correlated with the incidence of thyroid cancer, there was a significant increase in risk of thyroid cancer among those with lower serum selenium levels (odds ratio = 7.7; CI=1.3-44.7).

III. Strength of the Scientific Evidence

Below, the agency rates the strength of the total body of publicly available evidence.  The agency conducts this rating evaluation by considering the study type (e.g., intervention, prospective cohort, case-control, cross-sectional), the methodological quality rating previously assigned, the number of studies and number of subjects per group, whether the body of scientific evidence supports a health claim relationship for the U.S. population or target subgroup, whether study results supporting the proposed claim have been replicated,[70] and the overall consistency of the total body of evidence.[71]  Based on the totality of the scientific evidence, FDA determines whether such evidence is credible to support a qualified health claim for the substance/disease relationship and, if so, considers what qualifying language should be included to convey the limits on the level of scientific evidence supporting the relationship or to prevent the claim from being misleading in other ways.

Bladder and Urinary Tract Cancers

As discussed in section II, the evidence about a possible relationship between selenium intake and risk of bladder cancer is based on two articles reporting on a total of three nested case-control studies. Helzlsouer et al. (1989) reported that the observed risk for bladder cancer in men and women combined was not associated with blood selenium levels (Helzlsouer et al., 1989). Helzlsouer et al. (1989) did not analyze the association for each gender separately. The risk of getting bladder cancer is two to three times higher in men than in women.[72]  Michaud et al. (2005) observed no relationship between toenail selenium levels and bladder cancer risk in men, but did observe a negative relationship in women. However, the reported findings of Michaud et al. (2005) in women have not been replicated, and replication of scientific findings is important in order to substantiate results.[73]  Because there is only one observational study in support of a risk reduction claim in women, there is very little evidence from which to conclude that a risk reduction relationship actually exists.  Based on the above, FDA concludes that there is very limited credible evidence for a relationship between selenium supplements and reduced risk of bladder cancer in women. Therefore, FDA concludes that the existence of a relationship between selenium supplements and reduced risk of bladder cancer in women is highly uncertain. 

There is no credible evidence for a relationship between selenium supplements and reduced risk of bladder cancer in men. Furthermore, there were no intervention or observational studies from which scientific conclusions could be drawn about the relationship between selenium intake and risk of other cancers of the urinary tract. Therefore, FDA concludes that there is no credible evidence for a claim about selenium supplements and a reduced risk of bladder cancer in men or for a claim about selenium supplements and risk of other cancers of the urinary tract.  

Brain Cancer

As discussed in section II, there were no intervention or observational studies from which scientific conclusions could be drawn about the relationship between selenium intake and risk of brain cancer. Therefore, FDA concludes that there is no credible evidence for a claim about selenium supplements and reduced risk of brain cancer.

Breast Cancer

As discussed in section II, the evidence about a possible relationship between selenium intake and risk of breast cancer is based on six nested case-control studies.  None of the six nested case-control studies showed a significant relationship between blood or toenail selenium levels and risk of breast cancer (Criqui et al., 1991; Dorgan et al., 1998; Hunter et al., 1990a; Overvad et al., 1991; Ringstad et al., 1988; Willett et al., 1983).  Therefore, FDA concludes that there is no credible evidence for a claim about selenium supplements and reduced risk of breast cancer.

Colon and Digestive Tract Cancers

As discussed in section II, the evidence about a possible relationship between selenium intake and risk of colon and digestive tract cancers is based on three prospective observational studies on colorectal cancer (Peters et al., 2006; Schober et al. 1987; Wallace et al., 2003). None of the three observational studies on colorectal cancer observed a relationship between serum selenium and incidence of this cancer (Peters et al., 2006; Schober et al. 1987; Wallace et al., 2003). There were no intervention or observational studies from which scientific conclusions could be drawn about the relationship between selenium intake and risk of any other cancer of the digestive tract.  Therefore, FDA concludes that there is no credible evidence for a claim about selenium supplements and reduced risk of colon cancer or other cancers of the digestive tract.

Liver Cancer

As discussed in section II, there were no intervention or observational studies from which scientific conclusions could be drawn about the relationship between selenium intake and risk of liver cancer. Therefore, FDA concludes that there is no credible evidence for a claim about selenium supplements and reduced risk of liver cancer.

Lung and Respiratory Tract Cancer

As discussed in section II, the evidence about a possible relationship between selenium intake and risk of lung and respiratory tract cancers is based on four nested case-control lung cancer studies (Goodman et al., 2001; Menkes et al., 1986; Ringstad et al., 1988; Willett et al., 1983). None of these four studies observed a relationship between blood selenium levels and risk of lung cancer. There were no intervention or observational studies that evaluated the relationship between selenium and risk of any other respiratory tract cancer.  Therefore, FDA concludes that there is no credible evidence for a claim about selenium supplements and reduced risk of lung cancer or other respiratory tract cancers.

Prostate Cancer

As discussed in section II, the evidence about a possible relationship between selenium intake and risk of prostate cancer is based on one intervention study (Lippman et al., 2009) and eight nested case-control studies (Criqui et al., 1991; Brooks et al., 2001; Goodman et al., 2001; Li et al., 2004b; Nomura et al., 2000; Peters et al., 2007; Peters et al., 2008; Willett et al., 1983). 

Lippman et al. (2009) reported on the Selenium and Vitamin E Cancer Prevention Trial, a moderate-quality intervention study in which men over 50 were given a selenium supplement (200 µg/day of selenomethionine) (n=8,752) or a placebo (n=8,696).  The study found no significant difference in the five-year incidence of prostate cancer between subjects who received a daily selenium supplement (4.56%) and subjects who received a placebo (4.43%).

Six of the eight observational studies showed no beneficial relationship between blood selenium levels or selenium supplementation and the risk of prostate cancer (Criqui et al., 1991; Goodman et al., 2001; Li et al., 2004b; Peters et al., 2007; Willett et al., 1983). The six studies were all of moderate or high methodological quality. While Brooks et al. (2001), a moderate quality study, observed no significant difference in the average plasma selenium levels between the 52 prostate cancer cases and 96 controls, the study showed a significant relationship between higher serum selenium levels and lower risk of prostate cancer.  Nomura et al. (2000), a moderate quality study conducted on Japanese-American men, reported that there was a significant difference between the average serum selenium levels of 249 prostate cancer cases and 249 controls, as well as in the relationship between quartiles of serum selenium and risk of prostate cancer, with higher selenium levels being associated with lower cancer risk.  Consistency of findings among similar and different study designs is important for evaluating the strength of the scientific evidence.[74] Furthermore, the three largest studies (816, 724, and 586 cases, respectively) did not observe a relationship (Li et al., 2004b; Peters et al., 2007; Peters 2008).

Based on the above, FDA concludes that there is very limited credible evidence for a relationship between selenium supplements and reduced risk of prostate cancer. Although there are two observational studies supporting the relationship, FDA concludes that it is highly unlikely that selenium supplements reduce the risk of prostate cancer because a large intervention study and six of the eight observational studies found no evidence of such a relationship.  

Thyroid Cancer

As discussed in section II, the evidence about a possible relationship between selenium intake and risk of thyroid cancer is based on one nested case-control study (Glattre et al., 1989). This study showed that higher serum selenium levels were associated with reduced risk of thyroid cancer when the average serum selenium concentration of 43 thyroid cancer cases was compared to that of 129 controls. FDA finds that there is very limited credible evidence for a qualified health claim for selenium intake and a reduced risk of thyroid cancer.  However, the reported findings of Glattre et al. (1989) have not been replicated, and replication of scientific findings is important in order to substantiate results.[75]  Because there is only one observational study in support of the claim, there is very little evidence from which to conclude that a risk reduction relationship actually exists. 

Based on the above, FDA concludes that there is very limited credible evidence for a relationship between selenium supplements and reduced risk of thyroid cancer. Therefore, FDA concludes that the existence of a relationship between selenium supplements and reduced risk of thyroid cancer is highly uncertain. 

IV. Other Enforcement Discretion Factors

A qualified health claim on the label or in the labeling of selenium dietary supplements is required to meet all applicable statutory and regulatory requirements under the Act, with the exception of the requirement that a health claim meet the significant scientific agreement standard and the requirement that the claim be made in accordance with an authorizing regulation. 

A. Qualifying level of selenium in dietary supplements

The general requirements for health claims provide that, if the claim is about the effects of consuming the substance at other than decreased dietary levels, the level of the substance must be sufficiently high and in an appropriate form to justify the claim.  Where no definition for a    "high" level of the substance has been established by regulation, the claim must specify the daily dietary intake necessary to achieve the claimed effect (see 21 CFR 101.14(d)(2)(vii)).

FDA has defined a "high" level of selenium in 21 CFR 101.54(b), which provides that a food is "high" in a nutrient if it contains 20% or more of the Daily Value for the nutrient.  Thus, based on the current Daily Value for selenium in 21 CFR 101.9(c)(8)(iv), which is 70 micrograms (µg), dietary supplements that contain 14 µg or more selenium per reference amount customarily consumed are high in selenium.

Since a "high" definition is established for selenium, FDA intends to exercise enforcement discretion for dietary supplements bearing a qualified health claim about selenium and a reduced risk of bladder cancer, prostate cancer, or thyroid cancer when the dietary supplement contains selenium at a level that meets or exceeds the requirement for a high level of selenium as defined in 21 CFR 101.54(b) (i.e., 14 µg or more per reference amount customarily consumed, based on the current Daily Value for selenium).

B. Appropriate form of selenium in dietary supplementsAs discussed in section I.C., Safety Review, the use of selenium sulfide in dietary supplements is not safe and lawful under 21 CFR 101.14(b)(3) because this compound is reasonably anticipated to be a human carcinogen. Therefore, dietary supplements containing selenium sulfide are not eligible for a qualified health claim.

Claims for Reduced Risk of Prostate Cancer or Thyroid Cancer

All of the credible evidence to support qualified health claims for selenium and reduced risk of prostate cancer and thyroid cancer is based on studies that used serum or plasma selenium as a biomarker of intake.  As discussed above in Section II.D, Assessment of Observational Studies, the available scientific evidence suggests that serum and plasma selenium levels reflect intake of selenomethionine, but not intake of selenocysteine or inorganic forms of selenium (e.g., selenate and selenite).  Therefore, FDA intends to consider the exercise of its enforcement discretion for dietary supplements bearing a qualified health claim about selenium and a reduced risk of prostate cancer or thyroid cancer only when the dietary supplement contains selenomethionine. 

Claim for Reduced Risk of Bladder Cancer

The two studies that support a qualified health claim for selenium supplements and reduced risk of bladder cancer in women used toenail selenium as a biomarker of intake.  As discussed above in section II.D, the available scientific evidence suggests that toenail selenium levels reflect long-term intake of selenium in all forms.  Therefore, FDA intends to consider the exercise of its enforcement discretion for dietary supplements bearing a qualified health claim about selenium and a reduced risk of bladder cancer when the dietary supplement contains any form of selenium other than selenium sulfide.

V. Agency's Consideration of Disclaimers or Qualifying Language

FDA considered but rejected use of a disclaimer or qualifying language to accompany the proposed claims for which the agency found no credible evidence, i.e., for selenium supplements and reduced risk of urinary tract cancers other than bladder cancer, lung and respiratory tract cancers, colon and other digestive tract cancers, brain cancer, liver cancer, and breast cancer.  FDA concluded that neither a disclaimer nor qualifying language would suffice to prevent consumer deception in these instances, where there is no credible evidence to support the claims.  Adding a disclaimer or incorporating qualifying language that effectively characterizes the claim as baseless is not a viable regulatory alternative because neither the disclaimer nor the qualifying language can rectify the message conveyed by the unsubstantiated claim.  See, e.g., In re Warner-Lambert Co., 86 F.T.C. 1398, 1414 (1975), aff'd, 562 F.2d 749 (D.C. Cir. 1977) (pro forma statements of no absolute prevention followed by promises of fewer colds did not cure or correct the false message that Listerine will prevent colds); Novartis Consumer Health, Inc. v. Johnson & Johnson-Merck Consumer Pharms. Co., 290 F.3d 578, 598 (3d Cir. 2002) ("We do not believe that a disclaimer can rectify a product name that necessarily conveys a false message to the consumer."); Pearson v. Shalala, 164 F.3d 650, 659 (D.C. Cir. 1999) (the court stated that, where the weight of the evidence was against the claim, FDA could rationally conclude that the disclaimer "The FDA has determined that no evidence supports this claim" would not cure the misleadingness of a claim).  In such a situation, adding a disclaimer or qualifying language does not provide additional information to help consumer understanding but merely contradicts the claim.  Resort Car Rental System, Inc. v.  FTC, 518 F.2d 962, 964 (9th Cir.) (per curiam) (upholding FTC order to excise "Dollar a Day" trade name as deceptive because "by its nature [it] has a decisive connotation for which any qualifying language would result in contradiction in terms."), cert denied, 423 U.S. 827 (1975); Continental Wax Corp. v. FTC, 330 F.2d 475, 480 (2d Cir. 1964) (same); Pasadena Research Labs v. United States, 169 F.2d 375 (9th Cir. 1948) (discussing "self-contradictory labels").  In the FDA context, courts have repeatedly found such disclaimers ineffective.  See, e.g., United States v. Millpax, Inc., 313 F.2d 152, 154 & n.1 (7th Cir. 1963) (disclaimer stating that "no claim is made that the product cures anything, either by the writer or the manufacturer" was ineffective where testimonials in a magazine article promoted the product as a cancer cure); United States v. Kasz Enters., Inc., 855 F. Supp. 534, 543 (D.R.I.) ("The intent and effect of the FDCA in protecting consumers from . . . claims that have not been supported by competent scientific proof cannot be circumvented by linguistic game-playing."), judgment amended on other grounds, 862 F. Supp. 717 (1994).

VI. Conclusions

Based on FDA's consideration of the scientific evidence submitted with the petition and other pertinent scientific evidence, FDA concludes that there is no credible evidence to support qualified health claims for selenium dietary supplements and a reduced risk of urinary tract cancers other than bladder cancer, lung and other respiratory tract cancers, colon and other digestive tract cancers, brain cancer, liver cancer, or breast cancer. Thus, FDA is denying claims for these cancers.  However, FDA concludes that there is very limited credible evidence for qualified health claims for selenium dietary supplements and bladder cancer,  prostate cancer, and thyroid cancer, provided that the qualified claims are appropriately worded so as not to mislead consumers. 

The proposed claims state that the scientific evidence is "convincing but not yet conclusive" that selenium may reduce the risk of the cancers that are the subject of the proposed claims.  For the cancers where there was no credible evidence of a risk reduction relationship between selenium and the particular cancer, the proposed language "convincing but not yet conclusive" is clearly false.  Additionally, even for the cancers where there was  credible evidence of a risk reduction relationship between selenium and the particular cancer, the language "convincing but not yet conclusive" is false and misleading   Specifically, the phrase "convincing but not yet conclusive" in the proposed claims for bladder cancer, prostate cancer, and thyroid cancer is false and misleading because it suggests that there is a significant and consistent body of persuasive scientific evidence showing that selenium intake reduces the risk of the specified cancer, and that with a small amount of additional supporting research FDA may consider the evidence for the risk reduction effect to be conclusive.  As discussed in sections II and III of this letter, there are actually very few scientifically credible studies supporting a claim that selenium reduces the risk of bladder cancer, prostate cancer, or thyroid cancer.  Therefore, the agency believes that the language requested in the petition, "convincing but not yet conclusive," mischaracterizes the strength of the evidence and is likely to mislead consumers into believing that there is more scientific evidence showing that selenium reduces the risk of the specified cancers than actually exists. Instead, FDA intends to consider the exercise of its enforcement discretion for the following qualified health claims:

Bladder Cancer

"One study suggests that selenium intake may reduce the risk of bladder cancer in women.  However, one smaller study showed no reduction in risk.  Based on these studies, FDA concludes that it is highly uncertain that selenium supplements reduce the risk of bladder cancer in women."  

Prostate Cancer

"Two weak studies suggest that selenium intake may reduce the risk of prostate cancer. However, four stronger studies and three weak studies showed no reduction in risk.  Based on these studies, FDA concludes that it is highly unlikely that selenium supplements reduce the risk of prostate cancer."

Thyroid Cancer

 "One weak, small study suggests that selenium intake may reduce the risk of thyroid cancer. Based on this study, FDA concludes that it is highly uncertain that selenium supplements reduce the risk of thyroid cancer."

FDA intends to consider exercising its enforcement discretion for the above qualified health claims when all factors for enforcement discretion identified in Section IV of this letter are met.

Please note that scientific information is subject to change, as are consumer consumption patterns.  FDA intends to evaluate new information that becomes available to determine whether it necessitates a change in this decision.  For example, scientific evidence may become available that will support significant scientific agreement, that will support a qualified health claim for  one or more claims that were denied, that will no longer support the use of the above qualified health claims, or that may raise safety concerns about the substance that is the subject of the claims.

 

Sincerely,

Barbara O. Schneeman, Ph.D.
Director
Office of Nutrition, Labeling and Dietary Supplements
Center for Food Safety and Applied Nutrition


References

Allen NE, Morris JS, Ngwenyama RA, Key TJ. A case-control study of selenium in nails and prostate cancer risk in British men. Brit j Cancer 2004;90:1392-1396.

American Cancer Society, Cancer Facts and Figures, 2008.

American Institute for Cancer Research. Oesophagus. In: World Cancer Research Fund., American Institute for Cancer Research.: Food, Nutrition and the Prevention of Cancer: A Global Perspective. Washington, DC: The Institute, 1997, pp 118-129.

Beño I, Klvanová J, Magálová T, Brtková A. Blood levels of natural antioxidants in gatric and colorectal precancerous lesions and cancers in Slovakia. Neoplasma 2000;47:37-40.

Biomarkers Definitions Working Group. Biomarkers and surrogate endpoints: preferred definitions and conceptual framework. Clin Pharmacol Ther  2001;69(3):89-95.

Blot W, Li J, Taylor P, Guo W, Dawsey S, Wang G, Yang C, Zheng S, Gail M, Li G, Yu Y, Liu B, Tangrea J, SunY, Liu F, Fraumeni J, Zhang Y, Li B. Nutrition intervention trials in Linxian China: Supplementation with specific vitamin/mineral combinations, cancer incidence and disease specific mortality in the general population. J Natl Cancer Inst 1993;85:1483-1492.

Blot W, Li J, Taylor P, Guo W, Dawsey S, Bing L. The Linxian trials: mortality rates by vitamin-mineral intervention groups. Am J Clin Nutr 1995;62:1424S-1426S.

Bonelli L, Camoriano A, Ravelli P, Missale G, Bruzzi P, Aste H. Reduction of the incidence of metachronous ademonas of the large bowel by means of antioxidants. Proceedings of International Selenium Tellurium Development Association. Y. Palmieri ed. 1998. Scottsdale AZ pp 91-94.

Brooks JD, Metter EJ, Chan DW, Sokoll LJ, Landis P, Nelson WG, Muller D, Andres R, Carter HB. Plasma selenium level before diagnosis and the risk of prostate cancer development. J Urol 2001;166:2034-2038.

Burk RF, Norsworthy BK, Hill KE, Motley AK, Byrne DW. Effects of chemical form of selenium on plasma biomarkers in a high-dose supplementation trial. Cancer Epidemiol Biomarkers Prev 2006;15:804-810.

Burk RF, Hill KE, Motley AK. Plasma selenium in specific and non-specific forms. Biofactors 2001;14:107-114.

Burk RF, Levander OA. Selenium.  In:  Shils, ME, Olsen JA, Shike M, et al. eds. Modern Nutrition in Health and Disease, 10th Edition. Lea and Febriger, Baltimore, Maryland, 2006; Chapter 16. 

Butler JA, Thomsen C, Whanger PD, Robinson MF. Selenium distribution in blood fractions of New Zealand women taking organic and inorganic selenium. Am J Clin Nutr 1991 53;748-754.

Clark LC. The epidemiology of selenium and cancer. Fed Proc 1985;9:2584-2589.

Clark LC, Combs GF, Turnbull BW, Slate EH, Chalker DK, Chow J, Davis LS, Glover RA, Graham GF, Gross EG, Krongrad A, Lesher JL, Park HK, Sanders BB, Smith CL, Taylor JR. Effects of selenium supplementation for cancer prevention in patients with carcinoma of the skin. A randomized controlled trial. Nutritional Prevention of Cancer Study Group. JAMA 1996; 276:1957-1963.

Clark LC, Dalkin B, Krongrad A, Combs GF, Turnbull BW, Slate EH, Witherington R, Herlong JH, Janosko E, Carpenter D, Borosso C, Falk S, Rounder J. Decreased incidence of prostate cancer with selenium supplementation: results of a double-blind cancer prevention trial. Br J Urol 1998;81:730-734.

Coates RJ, Weiss NS, Daling JR, Morris JS, Labbe RF. Serum levels of selenium and retinol and the subsequent risk of cancer. Am J Epidemiol 1988;128:515-523.

Combs GF, Clark LC, Turnbull BW. Reduction of cancer risk with an oral supplement of selenium. Biomed Environ Sci 1997:10:227-234.

Combs GF. Food system-based approaches to improving micronutrient nutrition: The case for selenium. Biofactors 2000;12:39-43.

Criqui MH, Bangdiwala S, Goodman DS, Blaner WS, Morris S, Kritchevsky S, Lippel K, Mebane I, Tyroler HA. Selenium, retinol, retinol-binding protein, and uric acid associations with cancer mortality in a population-based prospective case-control study. Ann Epidemiol 1991;1:385-393.

Dawsey SM, Wang GQ, Taylor PR, Li JY, Blot WJ, Li B, Lewin KJ, Liu FS, Weinstein WM, Wiggett S. Effects of vitamin/mineral supplementation on the prevalence of histological dysplasia and early cancer of the esophagus and stomach: results from the Dysplasia Trial in Linxian, China. Cancer Epidemiol Biomarkers Prev 1994;3:167-172.

Dorgan JF, Sowell A, Swanson CA, Potischman N, Miller R, Schussler N, Stephenson HE. Relationships of serum carotenoids, retinol, alpha-tocopherol, and selenium with breast cancer risk: results from a prospective study in Columbia, Missouri (United States). Cancer Causes Control 1998;9:89-97.

Dresoti I. Selenium. J Food Nutr 1986;43:60-78.

Duffield-Lillico AJ, Reid ME, Turnbull BW, Combs GF, Slate EH, Fischbach LA, Marshall J.R, Clark LC. Baseline characteristics and the effect of selenium supplementation on cancer incidence in a randomized clinical trial: a summary report of the Nutritional Prevention of Cancer Trial. Cancer Epidemiol Biomarkers Prev 2002;11:630-639.

Duffield-Lillico AJ, Dalkin BL, Reid ME, Turnbull BW, Slate EH, Jacobs ET, Marshall JR, Clark LC. Selenium supplementation, baseline plasma selenium status and incidence of prostate cancer: an analysis of the complete treatment period of the Nutritional Prevention of Cancer Trial. BJU Int 2003a;91:608-612.

Duffield-Lillico AJ, Slate EH, Reid ME, Turnbull BW, Wilkins PA, Combs GF, Park HK, Gross EG, Graham GF, Stratton MS, Marshall JR, Clark LC. Selenium supplementation and secondary prevention of nonmelanoma skin cancer in a randomized trial. J Natl Cancer Inst 2003b;95:1477-1481.

Federal Judicial Center, Reference Manual on Scientific Evidence, Second Edition, 2000, page 93.

Fex G, Pettersson B, Akesson B. Low plasma selenium as a risk factor for cancer death in middle-aged men. Nutr Cancer 1987:10;221-229.

Finley J, Matthys L, Shuler T, Korynta E. Selenium content of foods purchased in North Dakota. Nutr Res 1996;16:723-728.

Flegal KM. Evaluating epidemiologic evidence of the effects of food and nutrient exposures. Am J Clin Nutr 1999;69:1399S-1344S.

Garland M, Morris JS, Stampfer MJ, Colditz GA, Spate VL, Baskett CK, Rosner B, Speizer Fem Willett WC, Hunter DJ. Prospective study of toenail selenium levels and cancer among women. J Natl Cancer Inst 1995;87:497-505.

Glattre E,  Thomassen Y, Thoresen SO, Haldorsen T, Lund-Larsen PG, Theodorsen L, Aaseth J. Prediagnostic serum selenium in a case-control study of thyroid cancer. Int J Epidemiol 1989;18:45-49.

Goodman GE, Schaffer S, Bankson DD, Hughes MP, Omenn GS. Predictors of serum selenium in cigarette smokers and the lack of association with lung and prostate cancer risk. Cancer Epidemiol Biomarkers Prev 2001;10:1069-1076.

Gromadzinska J, Wasowicz W, Rydzynski K, Szeszenia-Dabrowska N. Oxidative-stress markers in blood of lung cancer patients occupationally exposed to carcinogens. Biol Trace Elem Res 2003;91:203-215.

Hartman TJ, Albanes D, Pietinen P, Hartman AM, Rautalahti M, Tangrea JA, Taylor PR. The association between baseline vitamin E, selenium, and prostate cancer in the alpha-tocopherol, beta-carotene cancer prevention study. Cancer Epidemiol 1998;7:335-340.

Hartman TJ, Taylor PR, Alfthan G, Fagerstrom R, Virtamo J, Mark SD, Virtanen M, Barrett MJ, Albanes D. Toenail selenium concentration and lung cancer in male smokers (Finland). Cancer Causes and Control 2002;13:923-928.

Helzlsouer KJ, Comstock GW, Morris JS. Selenium, lycopene, alpha-tocopherol, beta-carotene, retinol, and subsequent bladder cancer. Cancer Res 1989;49:6144-6148.

Helzlsouer KJ,, Huang HY, Alberg AJ, Hoffman S, Burke A, Norkus EP, Morris JS, Comstock GW. Association between α-tocopherol, γ-tocopherol, selenium and subsequent prostate cancer. J Natl Cancer Inst 2000;92:2018-2023.

Hercberg S, Galan P, Preziosi P, Bertrais S, Mennen L, Malvy D, Roussel AM, Favier A, Briancon S. The SU.VI.MAX Study: a randomized, placebo-controlled trial of the health effects of antioxidant vitamins and minerals. Arch Intern Med 2004;164:2335-2342.

Hercberg S, Ezzedine K, Guinot C, Preziosi P, Galan P, Bertrais S, Estaquio C, Briancon S, Favier A, Latreille, Malvy D. Antioxidant supplementation increases the risk of skin cancers in women but not in men. J Nutr 2007;137:2098-2105.

Hietanen E, Bartsch Bérézait, Camus AM, McClinton S, Eremin O, Davidson L, Boyle P. Diet and oxidative stress in breast, colon and prostate cancer patients: a case-control study. Euro J Clin Nutr 1994;48:575-586.

Hoenjet KMJLF, Dagnelie PC, Delare KPJ, Wijckmans NEG, Zambon JV, Oosterhof GON. Effect of nutritional supplement containing vitamin E, selenium, vitamin C and coenzyme Q10 on serum PSA in patients with hormonally untreated carcinoma of the prostate: A randomized placebo-controlled study. Euro Urol 2005;47:433-440.

Hofstad B, Almendingen K, Vant M, Andersen SN, Owen RW, Larsen S, Osnes M. Growth and recurrence of colorectal polyps: a double-blind 3-year intervention with calcium and antioxidants. Digestion 1998;59:148-156.

Hord NG, Fenton JI. Context is everything: mining the normal and preneoplastic microenvironment for insights into the diet and cancer risk conundrum.  Mol Nutr Food Res 2007; 51:100-224.

Hunter DJ, Morris JS, Stampfer MJ, Colditz GA, Speizer FE, Willett WC. A prospective study of selenium status and breast cancer risk. JAMA 1990a;264:128-1131.

Hunter DJ, Morris JS, Chute CG, Kushner E, Colditz GA, Stmpfer MJ, Speizer FE, Willett WC.

Predictors of selenium concentration in human toenails. Am J Epidemiol 1990b;132:114-122.

Institute of Medicine (IOM), National Academy of Sciences.  Dietary Reference Intakes for Vitamin C, Vitamin E, Selenium, and Carotenoids.  Chapter 7, Selenium.  National Academy Press.  Washington, D.C. 2000.

Institute of Medicine (IOM), National Academy of Sciences. Dietary Supplements: A Framework for Evaluating Safety. Chapter 7, Categories of Scientific Evidence – In Vitro Data. National Academy Press. Washington, D.C. 2005.

Jain MG, Hislop GT, Howe GR, Ghadirian P. Plant foods, antioxidants, and prostate cancer risk: Findings from case-control studies in Canada. Nutrition and Cancer 1999;34:173-184.

Kabuto M, Imai H, Yonezawa C, Neriishi K, Akiba S, Kato H, Suzuki T, Land CE, Blot WJ. Prediagnostic serum selenium and zinc levels and subsequent risk of lung and stomach cancer in Japan. Cancer Epidemiol Biomarkers Prev 1994;3:465-469.

Kasperek K, Kiem J, Iyenger GV, Feinendegen LE. Concentration differences between serum and plasma of the elements cobalt, iron, mercury, rubidium, selenium and zinc determined by neutron activation analysis. Science Total Environment. 1981;17:133-143.

Key TJ, Silcocks PB, Davey GK, Appleby PN, Bishop DT. A case-control study of diet and prostate cancer. Brit J Cancer 1997;76(5):678-687.

Key TJ, Schatzkin A, Willett WC, Allen NE, Spencer EA, Travis RC.  Diet, nutrition and the prevention of cancer.  Public Health Nutrition 2004;7:187-200.

Knekt P, Aromaa A, Maatela J, Alfthan G, Aaran R, Teppo L, Hakama M.  Serum vitamin E, serum selenium and the risk of gastrointestinal cancer. Int  J Cancer 1988;42:846-850.

Knekt P, Aromaa A, Maatela J, Alfthan G, Aaran RK, Hakama M, Hakilinen T, Peto R, Teppo L. Serum selenium and subsequent risk of cancer among Finnish men and women. J Natl  Cancer Inst 1990;82:864-868. 

Knekt P, AromaaA, Maatela J, Alfthan G, Aaran R, Nikkari T, Hakam M, Hakulinen T, Teppo L. Serum micronutrients and risk of cancers of low incidence in Finland. Am J Epidemiol 1991;134:356-361.

Knekt P, Marniemi J, Teppo L, Heliovaara M, Aromaa A. Is low selenium status a risk factor for lung cancer? Am J Epidemiol 1998;148:975-982.

Kok F. J. Kok, A. M. de Bruijn, A. Hofman, R. Vermeeren, and H. A. Valkenburg. Is serum selenium a risk factor for cancer in men only? Am J Epidemiol 1987 125:12-16.

Kornitzer M, Valente F, De Bacquer D, Neve J, de Backer G. Serum selenium and cancer mortality: a nested case-control study within an age- and sex-stratified sample of the Belgian adult population. Eur J Clin Nutr 2004;58:98-104.

Krishnaswamy K, Prasad MP, Krishna TP, Annapurna VV, Reddy GA. A case study of nutrient intervention of oral precancerous lesions in India. Eur J Cancer B oral Oncol 1995;31B:41-48.

Kranse R, Dagnelie PC, van Kemenade MC, de Jong FH, Blom JHM, Tijburg LBM, Westrate JA, Schroder FH. Dietary intervention in prostate cancer patients: PSA response in a randomized double-blind placebo-controlled study. Int J Cancer 2005;113:835-840.

Krogh V, Pala V, Vinceti M, Berrino F, Ganzi A, Micheli A, Muti P, Vescovi L, Ferrari A, Fortini K, Sieri S, Viboli G. Toenail selenium as biomarker: reproducibility over one-year period and factors influencing reproducibility. J Trace Elem Med Biol 2003;17:31-36.

Kune G, Watson G. Colorectal cancer protective effects and the dietary micronutrients folate, methionine, vitamins B6, B12, C, E, selenium, and lycopene. Nutr Cancer 2006;56(1):11-21.

Levander OA. Considerations on the assessment of selenium status. Fed Proc 1985;44:2579-2583.

Li H, Li HQ, Wang Y, Xu HX, Fan WT, Wang ML, Sun PH, Xie XY. An intervention study to prevent gastric cancer by micro-selenium and large dose of allitridum. Chinese M J 2004a;117:1155-1160.

Li H, Stampfer MJ, Giovannucci EL, Morris JS, Willett WC, Gaziano JM, Ma J. A prospective study of plasma selenium levels and prostate cancer risk. J Natl Cancer Inst 2004b;96: 696-703.

Li JY, Taylor PR, Li B, Dawsey S, Wang GQ, Ershow AG, GuoW, Liu SF, Yang CS, Shen Q. Nutrition intervention trials in Linxian, China: multiple vitamin/mineral supplementation, cancer incidence, and disease-specific mortality among adults with esophageal dysplasia. J Natl Cancer Inst 1993;85:1492-1498.

Lichtenstein AH, Russell RM.  Essential Nutrients: Food or Supplements?  JAMA 2005;294:351-358.

Limburg PL, Wei W, Ahnen DJ, Qiao Y, Hawk ET, Wang G, Giffen CA, Wang G, Roth MJ, Lu N, Korn EL, Ma Y, Caldwell KL, Dong Z, Taylor PR, Dawsey SM. Randomzied, placebo-controlled, esophageal squamous cell cancer chemoprevention trial of selenomethionine and Celecoxib. Gastroenterol 2005;129:863-873.

Lipsky K, Zigeuner R, Zischka M, Schips L, Pummer K, Rehak P, Hubmer G. Selenium levels of patients with newly diagnosed prostate cancer compared with control group. Urology 2004;63(5):912-916.

Lipman SM, Goodman PJ, Klein EA, Parnes HL, Thompson IM, Kristal AR, Santella RM, Probstfield JL, Moinpour CM, Albanes D, Taylor PR, Minasian LM, Hoque A, Thomas SM, Crowley JJ, Gaziano JM, Stanford JL, Cook ED, Fleshner NE, Lieber MM, Walther PJ, Khuri FR, Karp DD, Schwartz GG, Ford LG, Coltman CA. J Nat Cancer Inst 2005;97(2):94-102.

Lippman et al. Effect of selenium and vitamin E on risk of prostate cancer and other cancers. JAMA 2009;301:39-51.

Longnecker MP, Taylor PR, Levander OA, Howe SM, Veillon C, McAdam PA, Patterson KY, Holden JM, Stampfer MJ, Morris JS, Willett WC. Selenium in diet, blood and toenails in relation to human health in a seleniferous area. Am J Clin Nutr 1991;53:1288-1294.

Longnecker MP, Stampfer MJ, Morris JS, Spate V, Baskett C, Mason M, Willett WC. A 1-yr trial on the effect of high-selenium bread on selenium concentration in blood and toenails. Am J Clin Nutr 1993;57:408-413.

Longnecker MP, Stram DO, Taylor PR, Levander OA, Howe M, Veillon C, McAdam PA, Patterson KY, Holden JM, Morris JS, Swanson CA, Willett WC. Use of selenium concentration in whole blood, serum, toenails, or urine as a surrogate measure of selenium intake. Epidemiology 1996;7:384-390.

Mark SD, Qiao YL, Dawsey SM, Wu YP, Katki H, Gunter EW, Fraumeni JF, Blot WJ, Dong ZW, Taylor PR. Prospective study of serum selenium levels and incident esophageal and gastric cancers. J Natl Cancer Inst 2000; 92:1753-1763.

Menkes MS, Comstock GW, Vuilleumier JP, Helsing KJ,  Rider AA, Brookmeyer R. Serum beta-carotene, vitamins A and E, selenium, and the risk of lung cancer. N Engl J Med  1986;315:1250-1254.

Meyer F, Verreault R. Erythrocyte selenium and breast cancer risk. Am J Epidemiol 1987;125(5):917-919.

Meyer F, Galan P, Douville P, Bairati I, KegleP, Bertrais S, Estaquio C, Hercberg S. Antioxidant vitamin and mineral supplementation and prostate cancer prevention in the SU.VI.MAX trial. Int J Cancer 2005;116:182-186.

Michaud DS, Hartman TJ, Taylor PR, Pietinen P, Alfthan G, Virtamo J, Albanes D. No association between toenail selenium levels and bladder cancer risk. Cancer Epidemiology, Biomarkers and Prevention 2002;11:1505-1506.

Milde D, Novák, Stuzka V, Vyslouzil K, Machácek J. Serum levels of selenium, manganese, copper, and iron in colorectal cancer patients. Biol Trace Elem Res 2001;79:107-114.

Milner JA. Diet and Cancer: Facts and Controversies.  Nutr Cancer 2006; 56-216-224.

Morris JS, Stampfer MJ, Willett W. Dietary selenium in humans, toenails as an indicator. Biol Trace Elem Res 1983;5:529-537.

Moser-Veillon PB, Mangels AR, Patterson KY, Veillon C. Utilization of two different chemical forms of selenium during lactation using stable isotope tracers :an example of speciation in nutrition. Analyst 1992;117:559-562.

National Cancer Institute, Dictionary of Cancer Terms, http://www.cancer.gov/Templates/db_alpha.aspx?CdrID=45333.

Nelson RL, Davis FG, Sutter E, Kikendall JW, Sobin LH, Milner JA, Bowen PE. Serum selenium and colonic neoplastic risk. Dis Colon Rectum 1995;38(12):1306-1310.

Niskar AS, Paschal DC, Kieszak SM, Flegal KM, Bowman B, Gunter EW, Pirkle JL, Rubin C, Sampson EJ, McGeehin M. Serum selenium levels in the US population: Third National Health and Nutrition Examination Survey, 1988-1994. Biol Trace Elem Res 2003;91:1-10.

Nomura A, Heilbrun LK, Morris JS, Stemmermann GN. Serum selenium and the risk of cancer, by specific sites: case-control analysis of prospective data. J Natl Cancer Inst 1987;79: 103-108.

Nomura AM, Lee J, Stemmermann GN, Combs GF. Serum selenium and subsequent risk of prostate cancer. Cancer Epidemiol Biomarkers Prev 2000; 9:883-887.

Omenn GS, Goodman GE, Thornquist MD, Balmes J, Cullen MR, Glass A, Keogh JP, Meyskens FL, Valanis B, Williams JH, Barnhart S, Hammer S.  Effects of a combination of beta carotene and vitamin A on lung cancer and cardiovascular disease.  New Eng J Med 1996;334:1150-1155.

Overvad K, Wang DY, Olsen J, Allen DS, Thorling EB, Bulbrook RD, Hayward JL. Selenium in human mammary carcinogenesis: a case-cohort study. Eur J Cancer 1991;27:900-902.

Peleg I, Morris S, Hames CG. Is serum selenium a risk factor of cancer? Medical Oncology and Tumor Pharmacotherapeutics 1985;2:157-163.

Peters U, Chatterjee N, Chruch TR, Mayo C, Sturup S, Foster CB, Schatzkin A, Hayes RB. High serum selenium and reduced risk of advanced colorectal adenoma in a colorectal cancer early detection program. Cancer Epidemiol Biomarkers Prev 2006;125:315-320.

Peters U, Foster CB, Chatterjee, N, Schatzkin A, Reding D, Andriole GL, Crawford ED, Sturup S, Chancok SJ, Hayes RB. Serum selenium and risk of prostate cancer: a nested case-control study. Am J Clin Nutr 2007;85:209-217.

Peters U, Littman AJ, Kristal AR, Patterson RE, Potter JD, White E. Vitamin E and selenium supplementation and risk of prostate cancer in the Vitamins and lifestyle (VITAL) study cohort. Cancer Causes Control 2008;19:75-87.

Peto, R., R. Doll, J.D. Buckley, M.B. Sporn.  Can dietary beta-carotene materially reduce human cancer rates?  Nature 1981;290:201-208.

Pawlowicz Z, Zacharia BA, Trafikowska U, Maciag A, Marchaluk E, Nowicki A. Blood selenium concentrations and glutathione peroxidase activities in patients with breast cancer and with advanced gastrointestinal cancer. J Trace Elem Electrolytes Health Dis 1991;5:275-277.

Prasad MP, Mukunda MA, Krishnaswamy K. Micronuleci and carcinogen DNA adducts as intermediate endpoints in nutrient intervention trial of precancerous lesions in the oral cavity. Eur J Cancer B Oral Oncol 1995;31B:155-159.

Qu CX, Kamanagar F, Fan JH, Yu B, Sun XD, Taylor PR, Chen BF, Abnet CC, Qiao YL, Mark SD, Dawsey SM. Chemoprevention of primary liver cancer: A randomized, double-blind trial in Linxian, China. J Natl Cancer Inst 2007;99:1240-1247.

Reid ME, Duffield-Lillico AJ, Garland L, Turnbull BW, Clark LC, Marshall JR. Selenium supplementation and lung cancer incidence: an update of the nutritional prevention of cancer trial. Cancer Epidemiol Biomarkers Prev 2002;11:1285-1291.

Reid ME, Duffield-Lillico AJ, Sunga A, Fakih M, Alberts DS, Marshall JR. Selenium supplementation and colorectal adenomas: an anlysis of the nutritional prevention of cancer trial. Int J Cancer 2006;118:1777-1781.

Ringstad J, Jacobsen BK, Tretli S, Thomassen,Y. Serum selenium concentration associated with risk of cancer. J Clin Pathol 1988;41:454-457.

Robinson MF, Thomson CD, Jenkinson CP, Luzhen G, Whanger PD. Long-term supplementation with selenate and selenomethionine: urinary excretion by New Zealand Women. Brit J Nutr 1997;77:551-563.

Rogers MA, Thomas DB, Davis S, Weiss NS, Vaughan TL, Nevissa AE. A case-control study of oral cancer and pre-diagnostic concentrations of selenium and zinc in nail tissue. Int J Cancer 1991;48:182-188.

Russo MW, Murray SC, Wurzelmann JI, Woosley JT, Sandler RS. Plasma selenium levels and the risk of colorectal adenomas. Nutr Cancer 1997;28(2):125-129.

Sakoda LC, Graubard BI, Evans AA, London WT, Lin WY, Shen FM, McGlynn KA. Toenail selenium and risk of hepatocellular carcinoma mortality in Hainen City, China. Int J Cancer 2005;115:618-624.

Salonen JT, Alfthan G, Huttunen JK, Puska P. Association between serum selenium and the risk of cancer. Am J Epidemiol 1984;120:342-349.

Salonen JT, Salonen R, Lappetelainen R, Maenpaa PH, Alfthan G, Puska P. Risk of cancer in relation to serum concentrations of selenium and vitamins A and E: Matched case-control analysis of prospective data. Br Med J 1985;290:417-420.

Schattner M, Shike M. Nutrition Support of the Patient with Cancer.  In:  Shils, ME, Olsen JA, Shike M, et al. eds. Modern Nutrition in Health and Disease, 10th Edition.  Lea and Febriger, Baltimore, Maryland, 2006; Chapter 83, page 1292.

Schatzkin A, Freedman LS, Dawsey SM, Lanza E.  Interpreting precursor studies:  what polyp trials tell us about large bowel cancer.  J Natl Cancer Inst 1994;86:1053-1057.

Schober SE, Comstock GW, Helsing KJ, Salkeld RM, Morris JS, Rider AA, Brookmeyer R. Serologic precursors of cancer.  I. Prediagnostic serum nutrients and colon cancer risk. Am J Epidemiol 1987;126:1033- 1041.

Scieszka M, Danch A, Machalski M, Drózdz M. Plasma selenium concentration in patients with stomach and colon cancer in the upper Silesia. Neoplasma 1997;44(6):395-397.

Sempos CT, Liu K, Ernst ND.  Food and nutrient exposures: what to consider when evaluating epidemiologic evidence.  Am J Clin Nutr 1999;69:1330S-1338S.

Singh P, Kapil U, Shukla NK, Deo SV, Dwivedi SN. Association between breast cancer and vitamin C, vitamin E, and selenium levels: results of a case-control study in India. Asian Pacific J Cancer Prev 2005;6:177-180.

Spilker, B. Guide to Clinical Studies. Raven Press, New York, New York, 1991.

Swanson CA, Longnecker MP, Veillon C, Howe M, Levander OA, Taylor PR, McAdam PA, Brown CC, Stampfer MJ, Willett WC. Selenium intake, age, gender, and smoking in relation to indices of selenium status of adults residing in a seleniferous area. Am J Clin Nutr 1990;52: 858-862.

Swanson CA, Patterson BH, Levander OA, Veillon C, Taylor PR, Helzlsouer K, McAdama PA, Zech LA. Human [74Se] selenomethionine metabolism:A kinetic model. Am J Clin Nutr 1991; 54:274-277.

Taylor PR, Li B, Dawsey SM, Li JY, Yang CS, Guo W, Blot WJ. Prevention of esophageal cancer: the nutrition intervention trials in Linxian, China. Linxian Nutrition Intervention Trials Study Group. Cancer Res. 1994;54: 2029s-2031s.

The Alpha-Tocopherol; Beta Carotene Cancer Prevention Study Group.  The effect of vitamin E and beta carotene on the incidence of lung cancer and other cancers in male smokers.  New Eng J Med 1994;330:1029-1035.

Thomsen CD, Robinson MF. Urinary and fecal excretions and absorption of a large supplement of selenium: superiority of selenate over selenite. Am J Clin Nutr 1986;44:659-663.

Thomson CD, Robinson MF, Butler JA, Whanger PD. Long-term supplementation with selenate and selenomethionine: selenium and glutathione peroxidase in blood components of New Zealand women. Br J Nutr 1993;69:577-588.

Thomson CD. Assessment of requirements for selenium and adequacy of selenium status: a review. Eur J Clin Nutr 2004;58:391-402.

Tominaga K, Saito Y, Mori K, Miyazawa N, Yokoi K, Koyama Y, Shimamura K, Imura J, Nagai M. Japan J Clin Oncol 1992;22(2):96-101.

Ujiie S, Kikuchi H. The relation between serum selenium value and cancer in Miyagi, Japan: 5-year follow up study. Tohoku J Exp Med 2002;196:99-109.

van den Brandt PA, Goldbohm RA, van't Veer P, Bode P, Dorant E, Hermus RJ, Sturmans F. A prospective cohort study on selenium status and the risk of lung cancer. Cancer Res 1993a;53:4860-4865.

van den Brandt PA, Goldbohm RA, van't Veer P, Bode P, Dorant E, Hermus RJJ, Surmans F. A prospective cohort study on toenail selenium levels and risk of gastrointestinal cancer. J Nat cancer Inst 1993b;85:224-229.

van den Brandt PA, Goldbohm RA, van't Veer P, Bode P, Dorant E, Hermus RJJ, Surmans F. Toenail selenium levels and risk of breast cancer. Am J Epidemiol 1994;140:20-26.

van den Brandt PA, Zeegers MPA, Bode P, Goldbohm RA. Toenail selenium levels and the subsequent risk of prostate cancer: a prospective cohort study. Cancer Epidemiol, Biomarkers Prevention 2003;12:866-871.

van Noord PAH, Collette HJ, Maas MJ, de Waard F. Selenium levels in nails of premenopausal breast cancer patients assessed prediagnostically in a cohort-nested case-referent study among women screened in the DOM project. Int J Epidemiol 1987;16:318-322.

van Noord PAH, Maas MJ, van der Tweel I, Collette C. Selenium and the risk of postmenopausal breast cancer in the DOM cohort. Breast Cancer Research Treatment 1993;25:11-19.

van't Veer P, Kok FJ, Brants HA, Ockhuizen T, Sturmans F, Hermus RJ. Dietary fat and the risk of breast cancer. Int J Epidemiol 1990;19(1):12-18.

van't Veer P, Strain JJ, Fernandez-Crehuet J, Martin BC, Thamm M, Kardinaal AFM, Kohlmeier L, Huttunen JK, Martin-Moreno JM, Kok FJ. Tissue antioxidants and postmenopausal breast cancer: The European community mulitcentre study on antioxidants, myocardial infarction, and cancer of the breast (EURAMIC). Cancer Epidemiol Biomarkers Prevention 1996;5:441-447.

Veillon C, Patterson KY, Button LN, Sytkowski AJ. Selenium utilization in humans – a long-term, self-labeling experiment with stable isotopes. Am J Clin Nutr 1990;52:155-158.

Vinceti M, Nacci G, Rocchi E, Cassinardi T, Vivoli R, Marchesi C, Bergomi M. Mortality in a population with long-term exposure to inorganic selenium via drinking water. J Clin Epidemiol 2000;53:1062-1068.

Virtamo J, Valkeila E, Alfthan G, Punsar S, Huttnen JK, Karvonen MJ. Serum selenium and risk of cancer. A prospective follow-up of nine years. Cancer 1987;60:145-148. 

Wallace K, Byers T, Morris JS, Cole BF, Greenberg ER, Baron JA, Gudino A, Spate V, Karagas MR. Prediagnostic serum selenium concentration and the risk of recurrent colorectal adenoma: A nested case-control study. Cancer Epidemiol Biomarkers Prev 2003;12:464-467.

Wang GQ, Dawsey SM, Li JY, Taylor PR, Li B, Blot WJ, Weinstein WM, Liu FS, Lewin KJ, Wang H. Effects of vitamin/mineral supplementation on the prevalence of histological dysplasia and early cancer of the esophagus and stomach: results from the General Population Trial in Linxian, China. Cancer Epidemiol Biomarkers Prev 1994;3:161-166.

Wei WQ, Abnet CC, Qiao YL, Dawsey SM, Dong ZW, Sun XD, Fan JH, Gunter EW, Taylor PR, Mark SD. Prospective study of serum selenium concentrations and esophageal and gastric cardia cancer, heart disease, stroke, and total death. Am J Clin Nutr 2004;79:80-85.

West DW, Slattery ML, Robison LM, French TK, Mahoney AW. Adult dietary intake and prostate cancer risk in Utah: a case-control study with special emphasis on aggressive tumors. Cancer Causes Control 1991;2:85-94.

Willett WC, Polk BF, Morris JS, Stampfer MJ, Pressel S, Rosner B, Taylor JO, Schneider K, Hames CG. Prediagnostic serum selenium and risk of cancer. Lancet 1983;2:130-134.

Willett WC. Overview of Nutritional Epidemiology. In Nutritional Epidemiology, (Oxford, United Kingdom: Oxford University Press), 1990, pp. 16-17.

Willett, W.C. (1998). Issues in analysis and presentation of dietary data. In Nutritional Epidemiology, Second Edition (Oxford, United Kingdom: Oxford University Press), pp. 339-340.

Wilson EB.  An Introduction to Scientific Research, pages 46-48, Dover Publishing, 1990.

Xia Y, Ha P, Hill K, Butler J, Whanger P. Distribution of selenium between fractions in erythrocytes, plasma, hair and fingernails of Chinese women living in selenium-deficient, -adequate, and –excessive areas of China. J Trace Elem Exp Med 2000;13:333-342.

Yamamoto T, Nishimori I, Tahara E.  Malignant tumors in atomic bomb survivors with special reference to the pathology of stomach and lung cancer. In: Shigematsu and A Kagan (eds) Cancer in Atomic Bomb Survivors pp 143-154. GANN Monograph on Cancer research No. 32 New York Plenum Publishing Co, 1986.

Yang CS, Sun Y, Yang QU, Miller,KW, Li GY, Zheng SF, Ershow AG, BlotWJ, Li JY. Vitamin A and other deficiencies in Linxian, a high esophageal cancer incidence area in northern China. J Natl Cancer Inst 1984;73:1449-1453.

Yang CS, Sun YH, Yang QP, Miller KW, Li GY, Zheng SF, Ershow AG, Li JY, Blot WJ. Nutritional status of the high esophageal cancer risk population in Linxian, People's Republic of China: effects of vitamin supplementation. Natl Cancer Inst Monogr 1985; 69:23-27.

Yang G, Zhou R, Yin S, Gu L, Yan B, Liu Y, Liu Y, Li X. Studies of safe maximal daily dietary selenium intake in a seleniferous area of in China. I. Selenium intake and tissue selenium levels of the inhabitants.  J Trace Elem Electrolytes Health Dis 1989;3:77-87.

Yang X, Tian Y, Ha P, Gu L. Determination of selenomethionine content in grain and in human blood. Wei Sheng Yen Chiku (J Hyg Res) 1997;26:113-116.

Yoshizawa K, Willett WC, Morris SJ, Stamfer MJ, Spiegelman D, Rimm EB, Giovannucci E. Study of prediagnostic selenium level in toenails and risk of advanced prostate cancer. J Natl Cancer Inst 1998;90:1219-1224.

You WC, Brown LM, Zhang L, Li JY, Jin ML, Chang YS, Ma JL, Pan KF, Liu WD, Hu Y, Crystal-Mansour S, Pee D, Blot WJ, Fraumeni JF, Xu GW, Gail MH. Randomized double-blind factorial trial of three treatments to reduce the prevalence of precancerous gastric lesions. J Natl Cancer Inst 2006;98:974-983.

Yu S, ChuY, Gong X, Hou C, Li W, Gong H, Xie J. Regional variation of cancer mortality and incidence and its relation to selenium levels in China. Biol Trace Elem Res 1985;7: 21-29.

Yu SY, Zhu YJ, Li WG, Huang QS, Huang CZ, Zhang QN, Hou C. A preliminary report on the intervention trials of primary liver cancer in high-risk populations with nutritional supplementation of selenium in China. Biol Trace Elem Res 1991;29:289-294.

Yu MW, Hornig IS, Chiang YC. Plasma selenium levels and the risk of hepatocellular carcinoma among men with chronic hepatitis virus infection. Am J Epidemiol 1999;150:367-374.

Yu SY, Zhu YJ, Li WG. Protective role of selenium against hepatitis B virus and primary liver cancer in Qidong. Biol Trace Elem Res 1997;56:117-124.

Zacharia BA, Marchaluk-Wisniewska E, Maciag A, Peplinski J, Skokowski J, Lambrecht W. Decreased selenium concentration and glutathione peroxidase activity in blood and increase of these parameters in malignant tissue of lung cancer patients. Lung 1997;175:321-332.

Zeegers MPA, Goldbohm PA, Bode P, van den Brandt PA. Prediagnostic toenail selenium and risk of bladder cancer. Cancer Epidemiol Biomarkers Prevention 2002;11:1292-1297.

Appendix 1

Please see Docket Nos. FDA-2008-Q-0323 and FDA-2007-N-0152 or the following studies and documents.  The parenthetical indicates the source of the study or document (i.e., Petition, Comment, FDA).  The studies or documents labeled with a (C) for Comment were received in response to a notice in the December 21, 2007 Federal Register (72 FR 72738; Docket No. FDA-2007-N-0152 (formerly 2007N-0464)) (reevaluation notice) announcing the agency's intent to reevaluate the scientific evidence for the existing qualified health claims for selenium and certain cancers (see Docket No. FDA-2002-P-0418 (formerly 2002P-0457)), along with three other health claims.  Although the reevaluation notice is not addressed in this letter, relevant studies and other documents supplied with comments to the reevaluation notice were considered in FDA's scientific evaluation of the qualified health claim petition to which this letter responds.

Articles written in a foreign language

Li et al., 2000 (P)

Animal Studies

Jamba et al., 1997 (P)

Medina et al., 1980 (P)

Rao et al., 2001 (P)

Schrauzer and Ishmael, 1974 (P)

Schrauzer et al., 1980a (P)

Schrauzer et al., 1978 (P)

Schrauzer et al., 1976a (P)

Schrauzer et al., 1976b (P)

Schrauzer et al., 1980b (P)

Schrauzer et al., 2006 (P)

Schrauzer et al., 2008 (P)

Waters et al., 2003 (P) 

In vitro Studies

Achanzar et al., 1986 (P)

Bhamre et al., 2002 (P)

Chun et al., 2007 (P)

Diwadkar-Navsariwala et al., 2006 (P)

Guimares et al., 1996 (P)

Ip and Lisk, 1997 (P)

Leinfelder et al., 1990 (P)

Roopari et al., 2007 (P)

Webber et al., 1985 (P)

Zhu et al. 1996 (P)

Zou et al., 2000 (P)

News releases/editorials/commentary

Fahey and Finley, 2004 (P)

Lippman et al., 2005 (F)

Shamberger and Frost 1969 (P)

Vinceti et al., 2003 (P)

Government documents

Bialostosky et al., 2002 (P)

Passwater and Olson, US patent 6.090.414 (P)

Taylor, (FDA) 2003 (P)

Federal Register, 72 FR 72738 (Dec. 21, 2007) (P)

www.cancer.gov/cancertopics/factsheet/prevention/Select, 2008 (P)

www.cancer.gov/clinicaltrials/ft-ECOG-5597, 2008 (P)

Book Chapter, Review Article or Meta-analysis

Brinkman et al., 2006 (P)

Burke and Hill, 1993 (P)

Cheung et al., 2008 (P)

Combs and Lu, 2001 (P)

Combs, 2005 (P)

Combs and Gray, 1998 (P)

Dadiminty and Gao, 2008 (C)

Dagnelie et al., 2004 (C)

Das et al., 2007 (P)

Diwadkar-Navsariwala and Diamond, 2004 (P)

Duntas 2006 (P)

Etminan et al.,2005 (C)

Greenwald et al., 2007 (P)

Hatfield and Gladyshev, 2002 (P)

Hesketh and Villette, 2002 (P)

Huang et al., 2006 (P)

Ip, 1998 (P)

Jacobs et al., 2004a (C)

Kiremidjjian-Schumacher et al., 1998 (P)

Klein, 2004 (C)

Knekt, 2001 (P)

Meplan et al., 1999 (P)

Naithani, 2008 (C)

Navarro et al., 2007 (C)

Ne've, 1995 (P)

Nelson, 1984 (P)

Rayman, 2005 (P)

Reilly et al., 2006 (P)

Schrauzer et al., 1977a (P)

Schrauzer et al., 1977b (P)

Schrauzer, 1977 (P)

Schrauzer, 1978 (P)

Schrauzer, 2003 (P)

Schrauzer, 2000 (P)

Spallholz, 2001 (P)

Stadtman, 1990 (P)

Van den Brandt and Goldbohm, 2006 (P)

Whanger, 2004 (P)

Willett and Stampfer, 1986 (P)

Zhuo et al., 2004 (P)

Studies that reported no data on the relationship between selenium intake and risk of  cancers for which the petition requested qualified health claims

Akbaraly et al., 2005 (P)

Black, 1947 (P)

Burney et al., 1989 (P)

Coughlin et al., 1996 (P)

Duffield-Lillico et al. 2003(b) (P)

Erikson et al., 1951 (P)

Faedo et al., 2004 (P)

Gianduzzo et al., 2003 (C)

Gromadzinska et al., 1996 (P)

Habib et al., 1976 (P)

Hercberg et al., 2007 (C)

Hickey et al., 2001 (P)

Karunasinghe et al., 2006 (C)

McElroy et al., 2006 (P)

Pagmantidis et al., 2008 (P)

Plackson et al., 2003 (P)

Rotruck et al., 1973 (P)

Sabichi et al., 2006 (C)

Schwarz and Foltz, 1957 (P)

Schopfer et al., 2004 (P)

Schrauzer and Rhead, 1971 (P)

Schrauzer et al., 1973 (P)

Siddiqui et al., 2006 (P)

Wright et al., 2004 (C)

Zhou et al., 2003 (P)

Intervention study that reported preliminary data without final results

Bonelli et al., 1998 (P)

Intervention study that did not evaluate the independent effect of selenium

Hofstad et al., 1998 (P)

Intervention studies conducted on subjects who had prostate cancer

Hoenjet et al., 2005 (C)

Kranse et al., 2005 (C)

Intervention studies that measured intermediate endpoints other than surrogate endpoints of cancer risk

Hoenjet et al., 2005 (C)

Joniau et al., 2007 (C)

Kranse et al., 2005 (C)

Krishnaswamy et al., 1995 (P)

Limburg et al., (2005) (F)

Prasad et al., 1995 (F)

Intervention studies that included post-hoc analysis of secondary cancers

Clark et al., 1996 (P)

Clark et al., 1998 (P)

Duffield-Lillico et al., 2002 (P)

Duffield-Lillico et al., 2003a (P)

Hercberg et al., 2004 (P)

Hercberg et al., 2007 (C)

Meyer et al., 2005 (P)

Reid et al., 2002 (P)

Reid et al., 2006 (C)

Study population not applicable to the general U.S. population (malnourished, selenium deficient or high prevalence of hepatitis or H. pylori)

Intervention Studies

Blot et al., 1993 (malnourished) (P)

Blot et al., 1995 F (malnourished) (F)

Dawsey et al., 1994 (malnourished) (F)

Li et al., 1993 (malnourished) (F)

Li et al., 2004a (H. pylori) (P)

Limburg et al., 2005 (malnourished) (F)

Qu et al., 2007 (malnourished and hepatitis) (C)

Taylor et al., 1994 (malnourished) (F)

Wang et al., 1994 (malnourished et al., 1994) (F)

Yu et al., 1991 (hepatitis B) (P)

Yu et al., 1997 (hepatitis B) (P)

You et al., 2006 (H. pylori)  (C)

Observational Studies

Hartman et al., 2002 (selenium deficient) (F)

Kabuto et al. 1994 (atomic bomb survivors) (P)

Knekt et al., 1988 (selenium deficient) (F)

Knekt et al., 1991 (selenium deficient) (F)

Knekt et al., 1990 (selenium deficient) (F)

Knekt et al., 1998 (selenium deficient) (P)

Mark et al., 2000 (selenium deficient) (F)

Michaud et al., 2002 (selenium deficient) (F)

Sakoda et al., 2005 (hepatitis)  (F)

Salonen et al., 1984 (selenium deficient) (F)

Salonen et al., 1985 (selenium deficient) (F)

Van den Brandt et al., 1994 (selenium deficient(P)

Van den Brandt et al., 1993a (selenium deficient) (P)

Van den Brandt et al., 1993b (selenium deficient) (P)

Van den Brandt et al., 2003 (selenium deficient) (P)

Van't Veer et al., 1996 (selenium deficient) (P)

Virtamo et al., 1987 (selenium deficient) (F)

Wei et al., 2004 (malnourished) (F)

Yu et al., 1999 (hepatitis) (P)

Zeegers et al., 2002 (selenium deficient) (P)

Studies that did not evaluate individual cancers

Intervention Studies

Blot et al., 1995 (F)

Hercberg et al, 2004 (P)

Observational Studies

Fex et al., 1987 (F)

Kok et al., 1987 (F)

Kornitzer et al., 2004 (F)

Observational studies that conducted inappropriate statistical analysis or  failed to adjust for confounders

Coates et al., 1988 (F)

Nomura et al., 1987 (F)

Peleg et al., 1985 (F)

Observational studies that did not adequately match or adjust for confounders

Helzlsouer et al., 2000 (P)

Garland et al., 1995 (P)

Van Noord et al., 1987 (F)

Van Noord et al., 1993 (F)

Yoshizawa et al., 1998 (P)

Observational studies that estimated selenium intake through dietary intake assessment methods or by known concentrations of selenium in drinking water

Cai et al., 2006 (C)

Clark et al., 1991 (P)

Hartman et al., 1998 (F)

Hietanen et al., 1994 (F)

Jain et al., 1999 (F)

Key et al., 1997 (F)

Kune and Watson, 2006 (F)

Lu et al., 2006 (P)

Mahabir et al., 2006 (P)

Shamberger and Willis, 1971 (P)

Vinceti et al., 2000 (water) (F)

West et al., 1991 (F)

Retrospective observational studies that measured a post-diagnostic biomarker of selenium intake in subjects with cancer

Allen et al., 2004 (F)

Beno et al., 2000 (F)

Charalabopoulos et al., 2006 (P)

Clark et al., 1993 (P)

Della Rovere et al., 2006 (P)

Fernandez-Banares et al., 2002 (P)

Gerhardsson et al., 1985 (P)

Ghadirian et al., 2000 (P)

Gromadzinska et al., 2003 (F)

Hardell et al., 1995 (P)

Jablonska et al., 2008 (P)

Jaskiewicz et al., 1988 (P)

Kellen et al., 2006 (P)

Lipsky et al., 2004 (F)

Mannisto et al., 2000 (P)

Meyer and Verrault, 1987 (F)

Milde et al., 2001 (F)

Miyamoto et al., 1987 (P)

Nelson et al., 1995 (F)

Pawlowicz et al., 1991 (F)

Reinhold et al., 1989 (P)

Rogers et al., 1991 (F)

Russo et al., 1997 (F)

Salonen et al., 1984 (F)

Salonen et al.  1985 (F)

Scieszka et al., 1997 (F)

Singh and Garg 1998 (P)

Singh et al., 2005 (F)

Tominaga et al., 1992 (F)

Ujiie and Kikuchi, 2002 (F)

Van't Veer et al., 1990 (F)

Van't Veer et al., 1996 (P)

Vogt et al., 2003 (P)

Westin et al., 1989 (P)

Yu et al., 1985 (F)

Zachara et al., 1997 (F)
 


Notes

[1] "Interim Procedures for Qualified Health Claims in the Labeling of Conventional Human Food and Human Dietary Supplements" (July 10, 2003). [http://www.fda.gov/Food/GuidanceComplianceRegulatoryInformation/GuidanceDocuments/FoodLabelingNutrition/ucm053832.htm]

[2] The April 28, 2003 decision letter is available on FDA's website at http://www.fda.gov/Food/LabelingNutrition/LabelClaims/QualifiedHealthClaims/ucm072780.htm.

[3] See Whitaker v. Thompson, 353 F.3d 947, 950-52 (D.C. Cir.) (upholding FDA's interpretation of what constitutes a health claim), cert. denied, 125 S. Ct. 310 (2004).

[4] See "Guidance for Industry: Evidence-based Review System for the Scientific Evaluation of Health Claims," January 2009.  [http://www.fda.gov/Food/GuidanceComplianceRegulatoryInformation/GuidanceDocuments/FoodLabelingNutrition/ucm073332.htm].

[5] For brevity, "disease" will be used as shorthand for "disease or health-related condition" in the rest of this section.

[6] In an intervention study, subjects similar to each other are randomly assigned to either receive the intervention or not to receive the intervention, whereas in an observational study, the subjects (or their medical records) are observed for a certain outcome (i.e., disease).  Intervention studies provide the strongest evidence for an effect.  See  supra, note 4.

[7] A meta-analysis is the process of systematically combining and evaluating the results of clinical trials that have been completed or terminated (Spilker, 1991).

[8] Review articles summarize the findings of individual studies.

[9] Other examples include book chapters, abstracts, letters to the editor, and committee reports.

[10] Certain meta-analyses may be used as part of the health-claim review process.  See supra, note 4.

[11] See supra, note 4 [Section III.F]. 

[12] Replication of scientific findings is important for evaluating the strength of scientific evidence (An Introduction to Scientific Research, E. Bright Wilson Jr., pages 46-48, Dover Publications, 1990).

[13] Consistency of findings among similar and different study designs is important for evaluating causation and the strength of scientific evidence (Hill A.B. The environment and disease: association or causation? Proc R Soc Med 1965;58:295-300); see also Systems to rate the scientific evidence, Agency for Healthcare Research and Quality [http://www.ahrq.gov/clinic/epcsums/strengthsum.htm#Contents], defining "consistency" as "the extent to which similar findings are reported using similar and different study designs."

[14] See supra, note 4 [Section III.F].

[15] National Cancer Institute, Dictionary of Cancer Terms [http://www.cancer.gov/dictionary/].

[16] National Cancer Institute, Dictionary of Cancer Terms [http://www.cancer.gov/dictionary/].

[17] See 21 U.S.C. § 321(n).

[18] FDA notes that its position on the use of non-specific language like "certain cancers" in health claims has changed over time.  The 2003 decision letter for the original selenium/cancer health claim petition stated the agency's intention to consider the exercise of enforcement discretion for a qualified health claim about reduced risk of "certain cancers" (see supra, note 2). In addition, regulations issued in 1993 as part of FDA's initial rulemaking to implement the health claim provisions of the Act authorize health claims for reduced risk of "some cancers" or "some types of cancer."  See 21 CFR 101.73, 101.76, 101.78.  As FDA's experience with health claims has grown, the agency's thinking on how to apply in practice the principle that each form of cancer is a different disease has evolved.  The evolution of the agency's thinking on this issue is discussed in the reevaluation notice (72 FR at 72740) and in FDA's guidance entitled "Evidence-Based Review System for the Scientific Evaluation of Health Claims," section III.B (see supra, note 4).  FDA believes that its current approach of naming in qualified health claim language the cancer(s) for which there is credible evidence of risk reduction best accords with the scientific understanding of cancer as a constellation of many diseases.   As discussed in the reevaluation notice (72 FR at 72740), the agency intends to consider, as part of its reevaluation of the 2003 selenium/cancer qualified health claims and the dietary lipids/cancer health claim in 21 CFR 101.73, replacing general claim language like "certain cancers" with references to individual named cancers.

[19] Stedman's Medical Dictionary, 28th ed. (2006).

[20] Stedman's Medical Dictionary, 28th ed. (2006).

[21] National Cancer Institute, Dictionary of Cancer Terms [http://www.cancer.gov/dictionary].

[22] American Cancer Society Glossary [http://www.cancer.org/docroot/GRY/GRY_0.asp?dictionary=&pagKey=C].  The Glossary contains no definition for "anticarcinogenic."

[23]See Whitaker v. Thompson, 353 F.3d at 950-52.

[24] FDA recognizes that its position on the use of "anticarcinogenic effects" in health claim language has changed since the 2003 decision letter for the original selenium/cancer health claim petition (see supra, note 2).  That letter said that the agency considered anticarcinogenic effects to be synonymous with cancer risk reduction and stated FDA's intention to consider the exercise of enforcement discretion for a qualified health claim about the production of anticarcinogenic effects in the body, based on studies that provided evidence of risk reduction for site-specific cancers.  As FDA's experience with health claims has grown, however, the agency's thinking on how to apply in practice the principle that each form of cancer is a different disease has evolved, as discussed in the reevaluation notice (72 FR at 72740) and in FDA's guidance entitled "Evidence-Based Review System for the Scientific Evaluation of Health Claims," section III.B (see supra, note 4).  In addition, since 2003 the agency has further considered the various possible meanings of the term "anticarcinogenic," based on definitions of other medical and scientific bodies.  FDA believes that its current approach of phrasing health claims in terms of risk reduction and naming in qualified health claim language the cancer(s) for which there is credible evidence of risk reduction best accords with the scope of health claims (see Whitaker, 353 F.3d at 950-52) and with the scientific understanding of cancer as a constellation of many diseases.  As part of FDA's reevaluation of the 2003 selenium/cancer qualified health claims, the agency intends to address these issues in the context of the "anticarcinogenic effects" claim reviewed in that letter.

[25]   The term "dietary ingredient" is defined in section 201(ff)(1) of the Act (21 U.S.C. § 321(ff)(1)) and includes vitamins; minerals; herbs and other botanicals; dietary substances for use by man to supplement the diet by increasing the total daily intake; and concentrates, metabolites, constituents, extracts, and combinations of the preceding types of ingredients. 

[26] Department of Health and Human Services, Public Health Service, National Toxicology Program.  The Report on Carcinogens, Eleventh Edition. 2005. [http://ntp.niehs.nih.gov/index.cfm?objectid=72016262-BDB7-CEBA-FA60E922B18C2540]

[27] An adenomatous polyp is a noncancerous polyp that starts in gland-like cells of the epithelial tissue (thin layer of tissue that covers organs, glands, and other structures within the body).  College of American Pathologists [http://www.cap.org/apps/docs/reference/myBiopsy/ColonAdenomatous.html]

[28] A polyp is a growth that protrudes from a mucous membrane.  National Cancer Institute, Dictionary of Cancer Terms [http://www.cancer.gov/dictionary/].

[29] Cancer that begins in the colon (the longest part of the large intestine) is called colon cancer, and cancer that begins in the rectum (the last several inches of the large intestine) is called rectal cancer.  Cancer that affects either of these organs may also be called colorectal cancer.  National Cancer Institute, Colorectal Cancer Prevention (Patient Version) [http://www.cancer.gov/cancertopics/pdq/prevention/colorectal/patient].  Colon and rectal cancers have the same risk factors and therefore many cancer studies measure these two cancers together.  National Cancer Institute [http://www.nlm.nih.gov/medlineplus/colorectalcancer.html].  Also, the clinical trials conducted by the NCI combine these two cancers together as colorectal cancer.  With respect to the proposed claims for colon and rectal cancer, the majority of the studies from which conclusions could be drawn did not differentiate between colon and rectal cancers.  Therefore, the agency evaluated colon and rectal cancer together.  Colorectal polyp recurrence has been used as a surrogate endpoint for colorectal cancer risk by the scientific community and has been used by NCI as a surrogate endpoint for colorectal cancer prevention (Schatzkin et al., 1994).

[30] This section generally discusses significant flaws in the reports of intervention studies from which scientific conclusions could not be drawn. Such studies may have other flaws in addition to those specifically mentioned, such as flaws related to the use of multi-nutrient dietary supplements in the intervention group.

[31] A reverse smoker of chutta is one who inserts the lit end of a rolled tobacco leaf into his or her mouth.

[32] Because the National Cancer Institute does not consider frequency of micronucleated cells and carcinogen DNA adducts as indicators of DNA damage to be risk factors for oral cancer, these endpoints cannot be surrogate endpoints of oral cancer risk.  See http://www.cancer.gov/cancertopics/pdq/prevention/oral/patient/page3(discussing risk factors for oral cancer).

[33]National Cancer Institute, Fact Sheet:  Prostate-Specific Antigen (PSA) Test [http://www.cancer.gov/cancertopics/factsheet/Detection/PSA]

[34] Abnormal squamous cells located in the esophagus.  Squamous cells are thin, flat cells that are found in the tissue that forms the surface of the skin, the lining of the hollow organs of the body, and the passages of the respiratory and digestive tracts.  National Cancer Institute, Dictionary of Cancer Terms [http://www.cancer.gov/dictionary]

[35] The National Cancer Institute considers acid reflux to be a risk factor for esophageal cancer. http://www.cancer.gov/cancertopics/wyntk/esophagus/page5.  Acid reflux over the course of many years may cause abnormal squamous cells in the esophagus (squamous dysplasia).

[36] The National Cancer Institute does not consider squamous dysplasia to be a risk factor for esophageal cancer and therefore squamous dysplasia cannot be a surrogate endpoint for risk of esophageal cancer.  See http://www.cancer.gov/cancertopics/wyntk/esophagus/page5 for the National Cancer Institute's list of risk factors for esophageal cancer.

[37] See supra, note 4 [Section III.C].

[38] The precise etiology of gastric and esophageal cancer is unknown.  However, two factors, high salt intake and Helicobacter pylori (H. pylori) infection, are associated with an increased risk of the disease and are external risk factors for gastric cancer.  High salt intake and H. Pylori infection are more prevalent in China and Japan than in the United States (Hoenberger et al., 2003; Key et al., 2004). 

[39] See supra, note 4 [Section III.D].

[40] See supra, note 4 [Section III.D].

[41] External causal factors are environmental, lifestyle, nutritional or cultural factors (e.g. smoking, chemical exposure, radiation, dietary factors, socioeconomic factors, and viruses).

[42] Internal causal factors are genetic, gender, race or inherent factors (e.g., metabolism and pH).

[43] Cancer Prevention and Control, Chapter 6, page 83, edited by Greenwald P., Kramer B., Weed D. Marcel Dekker Publishing, 1995.

[44] National Cancer Institute, Stomach (Gastric) Cancer Prevention (Patient Version) [http://www.nci.nih.gov/cancertopics/pdq/prevention/gastric/HealthProfessional/page1].

[45]Centers for Disease Control and Prevention, Helicobacter pylori Fact Sheet for Health Care Providers, July 1998 [http://www.cdc.gov/ulcer/files/hpfacts.PDF].

[46] Bias is defined as the result of a systematic error in the design or conduct of a study. There are many forms of bias that can influence the interpretation of data (Spilker, B. Guide to Clinical Studies. Pages 612-636, Raven Press, 1991).  As a consequence of bias, the observed results of a study may be different from the true results.

[47] Kavanaugh C, Trumbo P, Ellwood K. Qualified Health Claims for Calcium and Colorectal, Breast, and Prostate Cancers:  The U.S. Food and Drug Administration's Evidence-Based Review. Nutrition and Cancer 2009;61:157-164.

[48] Qualified Health Claims:  Letter Regarding Antioxidant Vitamins C and E and Reduction in the Risk of Site-Specific Cancers (Docket No. FDA-2008-Q-0299) [http://www.fda.gov/Food/LabelingNutrition/LabelClaims/QualifiedHealthClaims/ucm072756.htm].

[49] See supra, note 4 [Section III.D].

[50] The petition did not request a qualified health claim for skin cancer.

[51] http://www.cancer.gov/cancertopics/wyntk/prostate/page6. Screening for elevated PSA  and a digital rectal examination is used to detect a cancer or a less serious problem in the prostate.  (see http://www.cancer.gov/cancertopics/factsheet/Detection/PSA). 

[52] Plasma and serum are both liquid forms of the blood. There is no significant difference between serum and plasma selenium concentrations from the same subject (Kasperek et al., 1981). Therefore, plasma and serum selenium are used interchangeably for assessing selenium intake and status.

[53] Correlation is evaluated using a correlation coefficient (r). Correlation coefficients range from -1 (negative correlation) through +1 (positive correlation). The closer to 1, the stronger the correlation; the closer to zero, the weaker the correlation.

[54] As previously noted, absorption of selenium is relatively high from both conventional foods and chemically pure forms of selenium, such as those used in dietary supplements (Finley et al., 2006). 

[55] In a duplicate-plate food collection study, study subjects save duplicate portions of all foods and beverages consumed. These foods are then analyzed for selenium concentration. Collecting and analyzing the actual foods consumed by the study subjects avoids the limitations associated with the dietary assessment methods for selenium intake described above under "Dietary Selenium."

[56] National Cancer Institute, What You Need to Know About  Liver Cancer:  Risk Factors [http://www.cancer.gov/cancertopics/wyntk/liver/page4].

[57] Centers for Disease Control and Prevention, Prevention of Specific Infectious Diseases:  Hepatitis, Viral, Type B  [http://wwwn.cdc.gov/travel/yellowBookCh4-HepB.aspx].  From Health Information for International Travel 2008 (The Yellow Book), Chapter 4.

[58] A nested-case control study uses subjects from a defined cohort.  Cases are subjects diagnosed with the disease (i.e., cancer) in the cohort and controls are subjects selected from individuals at risk each time a case (i.e., cancer) is diagnosed. (Epidemiology: Beyond the Basics, page 34, Aspen Publishers, 2000).

[59] See supra, note 4 [Section III.E].

[60] See supra, note 4[Section III.B].

[61] Confidence intervals are ranges that provide a statistical analysis of comparative measures of risk (e.g., relative risk, odds ratio and hazard ratio).  Confidence intervals are significant when the entire range is less than or greater than "1" (e.g., 0.7-0.9 or 1.1-1.5). If the confidence interval includes "1", then it can be concluded that a relationship does not exist between the substance and the disease.

[62] See supra, note 4 [Section III.E].

[63] See supra, note 4 [Section III.D].

[64] See supra, note 4 [Sections III.D and III.E].

[65] P is a measure of the statistical significance of the linear relationship between the substance and risk of the disease.   For the outcome of an observational study to demonstrate a statistically significant difference in plasma, serum or toenail selenium levels between the cases and controls, p must be < 0.05.

[66] Odds ratio is the odds of developing the disease in exposed compared to unexposed individuals (Epidemiology: Beyond the Basics, page 29, Aspen Publishers, 2000). Odds ratio is a measure of risk and is calculated in case-control studies by measuring development of a disease (here, a specific type of cancer) in subjects based on exposure to the substance of interest (here, selenium).

[67] Relative risk is expressed as the ratio of the risk (e.g., incidence of the disease) in exposed individuals to that in unexposed individuals (Epidemiology: Beyond the Basics, page 93, Aspen Publishers, 2000).

[68] Gastrointestinal tract cancers are cancers of the digestive tract, which includes the mouth, esophagus, stomach, small and large intestines, rectum, and anus. National Cancer Institute, Dictionary of Cancer Terms [http://www.cancer.gov/dictionary].

[69]The hazard ratio is an estimate of relative risk. It describes the relative risk of the disease based on comparison of event rates.

[70] See supra, note 12.

[71] See supra, note 13.

[72] http://www.cancer.gov/cancertopics/wyntk/bladder/page4.

[73] See supra, note 12.

[74] See supra, note 13.

[75] See supra, note 12.