• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Determining the Strength of the Relationship Between a Food, Food Component, or Dietary Supplement Ingredient and Reduced Risk of a Disease or Health-Related Condition1

Posted on: Friday, 18 March 2005, 03:00 CST

ABSTRACT

Evaluation of the strength of the science between a food/ substance and a disease is not unique to the FDA and their review of petitions for health claims, although this is the focus of the symposium. Determining the strength of the science linking a food/ substance to a reduced risk of disease is also an important part of the process used to set the Dietary Reference Intake values (DRIs) and the Dietary Guidelines for Americans. It involves using an evidence-based system. An appreciation of how evidence-based systems work and how research studies attempting to show the relationship between a food/substance and a disease are evaluated should lead to the design of better studies. Better studies, in turn, will result in multiple benefits. J. Nutr. 135: 340-342, 2005.

KEY WORDS: * substance/disease relationships * evidence-based systems * health claims

Although the focus of this symposium is on evaluating the strength of the science between a food/substance and decreased risk of a disease or health-related condition in support of a health claim (which is regulated by the FDA), determining the strength of diet/disease relationships is not unique to the FDA. The Institute of Medicine, National Academy of Sciences, evaluated the strength of the science between nutrients and disease or health-related conditions in setting the Dietary Reference Intakes (DRIs)3 (1). The Dietary Guidelines Advisory Committees evaluate the strength of the science linking foods and nutrients to reduced risk of disease to make informed recommendations on guidelines for Americans (2-4). The American Dietetic Association evaluates the strength of the science linking a diet recommendation to an improvement in health (5,6). The overall goal of this paper is to show how the strength of the science between a food/nutrient/substance and a disease or health- related condition can be determined using an evidence-based review system. The focus is on the evaluation of health claims as a specific use of an evidence-based system.

The need to assess the strength of the relationship between a nutrient and a disease

National nutrition policy requires an ongoing assessment of how food/nutrient intake affects human health. Such information is needed to establish DRIs, the Dietary Guidelines, and to determine what health claims may be put on food labels. In addition, a variety of professional organizations with a focus on diet/disease relationships (e.g., the American Dietetic Association, the American Heart Association, the American Diabetes Association) evaluate these relationships to establish best practice guidelines for their members. The NIH may determine the strength of these diet/disease relationships in considering whether to support major clinical intervention trials testing the effect of a food/substance on decreased risk of a disease. A few specific examples follow.

The Dietary Reference Intake process. The DRI process expands on and replaces the series of reports called Recommended Dietary Allowances (RDAs) published in the United States, and Recommended Nutrient Intakes (RNIs) in Canada (1). The major differences between the DRI reports and the previous RDA reports are that DRI values can be established with decreased risk of chronic disease as an endpoint rather than the previous endpoint of protecting against deficiency diseases [see the original concept paper for establishing the DRIs (7)]. In addition, food substances not traditionally considered to be nutrients can be reviewed for possible assignment of a DRI value if sufficient data exist to warrant the evaluation (7). This means that for establishing each DRI value, an assessment must be made of the strength of the relationship between intake of the food/ substance and decreased risk of chronic disease. If strong data exist, it would be possible to assign an EAR value (Estimated Average Requirement). The EAR can then be used to calculate an RDA. For example, an EAR and an RDA exist for protein, based on the amount required to maintain nitrogen balance (8). If sufficient evidence exists but it is not as strong as that required to establish an EAR, then an AI (adequate intake) value might be assigned. An example of such an assignment is the AI for fiber, which is based on decreased risk of coronary heart disease (9). If insufficient evidence exists to establish either an. EAR or an AI, then no DRI value for intake is assigned. Thus, the DRI process is used to evaluate the strength of the relationship between a nutrient or other substance and decreased risk of disease.

The process of establishing the Dietary Guidelines. The Dietary Guidelines for Americans are the cornerstone of federal nutrition policy and nutrition education activities. They are issued jointly by the USDA and the Department of Health and Human Services and are updated every 5 y. The National Nutrition Monitoring and Related Research Act of 1990 requires the Secretaries of Agriculture and Health and Human Services to publish jointly every 5 y a report entitled, the Dietary Guidelines for Americans (10). The report must contain nutritional and dietary information and guidelines for the general public and be based on the preponderance of scientific and medical knowledge current at the time of publication (10). To determine what the guidelines should be, the guidelines advisory committee must evaluate the strength of the relationship between a food or nutrient and decreased risk of a disease and/or the promotion of human health. For example, the strength of the relationship of whole grain intake to reduced risk of heart disease and/or diabetes may be weighed to determine whether whole grains should be recommended over refined grains. Assigning an actual value for intake recommendations requires dose/response data.

Professional societies and their evaluation of diet/disease relationships. A number of professional societies issue practice guidelines based on an analysis of the strength of the relationship between diet and disease. For example, the American Dietetic Association scientifically evaluates the diet/disease relationship in support of their evidence-based dietetics practice (5,6). Dietitians can consult the evidence-based reviews to evaluate the strength of the relationship between a particular diet intervention and an outcome that will be beneficial to health (e.g., lowering cholesterol, modulating blood glucose levels) (5,6). Based on the strength of the relationship they can strongly recommend a diet intervention or, if the data are less compelling, suggest but not strongly recommend a diet intervention. Similarly, the American Diabetes Association uses an evidence-based review to determine its clinical practice guidelines (11).

The FDA and their evaluation of health claims. The health claims that are authorized by the FDA and appear on food labels are about the relationship between a substance and a disease. They were authorized in 1990 by the Nutrition Labeling and Education Act (12) and apply to both conventional foods and dietary supplements. The highest level of evidence behind a food/substance and reduced risk of a disease would meet the Significant Scientific Agreement (SSA) standard (13). More recently, a Task Force, appointed by Mark B. McClellan, Commissioner of Food and Drugs, was mandated to determine the best way to evaluate the strength of the science behind potential health claims for which the science was not as strong as that required for SSA health claims (14). Their report, issued in July 2003, is the basis for much of the information in the following section (14).

Evidence-based systems: the common mechanism for evaluating the strength of the relationship between a food/ substance and the risk of a disease

The common mechanism for evaluating the strength of the relationship between a food/substance and the risk of a disease is the use of an evidence-based rating system, which is a science- based systematic evaluation of the strength of the evidence behind a statement. In the case of the DRI process, the statement will recommend the amount of a nutrient that should he consumed to promote health; the amount will he based on a particular endpoint described and evaluated in the text. In the case of the Dietary Guidelines, the statement will be a guideline to help Americans eat healthier diets, with scientific documentation for the guidelines provided in the Guidelines Advisory Committee Report. In the case of health claims, the statement would rate the strength of the evidence behind a proposed substance/disease relationship, for use on the food label. Although evidence-based reviews differ from each other, most have the same essential components. For an excellent analysis of existing evidence-based systems and their evaluation see West et al. (15).

Applying an evidence-based system to qualified health claims

As noted above, if the strength of the relationship between a substance and a disease is very high, then it will meet the SSA standard for health claims (12,13). However, if the science behind the claim is not well enough established to meet the SSA standard, then qualifying language is required for the food label, and the claim is termed a "q\ualified health claim." The type of qualifying language will depend on the strength of the relationship; that is, the weaker the relationship, the stronger the qualifying language. It therefore becomes necessary to have a method to determine the strength of the relationship. The Food and Drug Administration uses an evidence-based system to rank the strength of the science so that the appropriate qualifying language can be used for a health claim. A general outline of their evidence-based approach can be found in FDA guidelines (16) and is discussed below.

Clarification of the question/statement to be evaluated. The first general step in the evaluation of a potential health claim is a critical look at the proposed claim statement itself. This usually takes the form of "Substance 'X' reduces the risk of disease 'Y' in (name the population)." This means that the substance has to be clearly defined (17), the disease has to be defined, and the target population must be defined. This is important because it determines which studies are relevant and which are not in support of the claim. If the proposed health claim was for a substance reducing the risk of prostate cancer in older men, then the most pertinent studies would test the substance itself in older men and determine whether there was a reduced incidence of their acquiring prostate cancer. Studies that used a mixture of substance "X" with other potentially active substances would not be rated as high because it would be difficult to separate out the effects of the subject of the claim from other substances. If the substance were tested in men who already had prostate cancer, the study would not be as highly rated because the claim is to reduce the risk of acquiring the disease not ameliorating the disease in people who already have it. Amelioration of a disease is considered to be a drug effect, not the effect of a food. If the substance were tested in adolescent males rather than older men it would not be as relevant.

Collection of all relevant studies. A study that would be considered of primary importance in support of a health claim would be a human study done in a nondiseased population. The study would be designed to determine directly the relationship between the subject of the proposed claim and the risk reduction of the disease itself or an accepted surrogate marker for the disease or health related-condition; it would use accepted study designs and statistical methods to evaluate the data. Human studies always rate higher than animal or in vitro studies. Primary data are rated higher than review articles or meta-analyses. Although nonhuman studies can be used as background information, they would not be considered primary studies in support of a health claim. As noted above, the study must he conducted in a nondiseased population and cannot be used to mitigate an existing disease. If the claim is for substance "X," then the studies with the greatest relevance would test substance "X," not a food that contains substance "X." For example, if the claim is for n-3 fatty acids, then the substance to be tested is n-3 fatty acids, not fish. A mixture or cocktail containing substance "X" and other substances is not a direct test, nor is an extract containing other bioactive components along with substance "X."

Evaluation of the quality of a study. The ranking of various study designs is based on minimizing bias, with randomized clinical trials receiving the highest score, followed by prospective cohort studies. Case-control studies and cross-sectional studies are ranked lower than either randomized clinical trials or prospective cohort studies, if one is not testing the relationship between the substance and risk reduction of the disease, then an accepted surrogate marker is required. This marker must be generally accepted as a surrogate marker, particularly with the FDA and NIH. For example, decreasing LDL cholesterol is considered an appropriate marker for reducing the risk of coronary heart disease (CHD), but other surrogate markers for decreased risk of CHD such as reductions in C-reactive protein or homocysteine levels may not be. Appropriate statistical tests are important, and inappropriate use of statistical methods can downgrade a study. A case in point is the use of paired t tests for multiple endpoints. For epidemiologic studies, known confounders have to be taken into consideration. Selection of the study population is very important, particularly that the population be healthy people who do not yet have the disease of interest. Studies conducted in populations that differ from the U.S. population in important ways (e.g., evidence of malnutrition or very different rates of disease) are not as relevant as populations similar to those in the United States. Before and after studies in 1 group are not valued as highly as a control and intervention group because the former do not account for external factors over time. Similarly, key differences between the control and test groups with respect to gender, BMI, or smoking, for example, would downgrade a study. The intervention in a clinical trial must be of sufficient length to establish the relationship, and the food or substance provided for the intervention must be adequately characterized. In addition, if the intervention affects the rest of the diet (e.g., substitution of 1 type of fat for a different type of fat, or fat at the expense of carbohydrate) then the rest of the diet must also be characterized. Reporting on how the intervention diet affected energy intake and weight gain is important because body weight may be an independent risk factor for certain diseases.

Rating of the strength of the entire body of evidence. In an evidence-based system, each primary study becomes part of the overall rating of the strength of the evidence. The overall quality of the studies combined with the quantity (total number of subjects), relevance to the U.S. population, and benefit and consistency of the findings are all taken into consideration. The end result of using the evidence-based ranking system is a statement linking a substance to a disease/health-related condition with a ranking related to the scientific evidence behind that statement. There is a clear and transparent demonstration of which research studies were evaluated to provide the ranking, and evidence tables showing the rigor of the evaluation. Trained scientists should come to similar conclusions using the same database. It should be noted, however, that science is an evolving process, and the strength of the evidence must be reevaluated over time. A review of how well diet/disease associations stand the test of time can be found in the Institute of Medicine publication (18).

Summary and conclusions

National nutrition policy requires the constant evaluation and reevaluation of the strength of the relationship between a food/ substance and reduced risk of a disease. There is increasing use of evidence-based systems to determine the strength of this relationship and such systems are used by the National Academy of Science for determining Dietary Reference Intakes; the Dietary Guidelines Advisory Committee for making recommendations on the guidelines for Americans; professional societies; and the FDA for evaluating health claims. With respect to health claims, the objective is that by using and refining an evidence-based system to clarify the strength of the science behind the claim, consumers will be able to make healthier food choices based on sound science.

ACKNOWLEDGMENTS

The author acknowledges the input of the following individuals at the FDA (CFSAN) for their contribution in helping to frame the original evidence-based guideline for qualified health claims: Kathleen Ellwood, Claudine Cavanaugh, Craig Rowlands, Christine Taylor, Paula Trumbo, and Elizabeth Yetley.

1 Presented as part of the symposium, "Nutrient Disease Relationships: Closing the Scientific Knowledge Gap," given at the 2004 Experimental Biology meeting on April 19, 2004, Washington, DC. The symposium was sponsored by the American Society for Nutritional Sciences and supported in part by McNeil Nutritionals. The proceedings are published as a supplement to The Journal of Nutrition. This supplement is the responsibility of the Guest Editors to whom the Editor of The Journal of Nutrition has delegated supervision of both technical conformity to the published regulations of The Journal of Nutrition and general oversight of the scientific merit of each article. The opinions expressed in this publication are those of the authors and are not attributable to the sponsors or the publisher, editor, or editorial board of The Journal of Nutrition. The Guest Editors for the symposium publication are Leila G. Saldanha, Scientific Consultant, Alexandria, VA, and Mary Ann Johnson, Department of Foods and Nutrition, University of Georgia, Athens, GA.

3 Abbreviations used: AI, adequate intake; CHD, coronary heart disease; DRI, Dietary Reference Intake; EAR, Estimated Average Requirement; RDA, Recommended Dietary Allowance; RNI, Recommended Nutrient Intake; SSA, Significant Scientific Agreement.

LITERATURE CITED

1. Institute of Medicine (2002) Introduction to dietary reference intakes. In: Dietary Reference Intakes: Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids, pp. 1.1- 1.14. The National Academies Press, Washington, DC.

2. McMurry, K. Y. (2003) Setting dietary guidelines: the US process. J. Am. Diet. Assoc. 103: S11-S16.

3. Dietary Guidelines Advisory Committee (1990) Report. U.S. Department of Agriculture, Washington, DC.

4. Cooper, M. J. & Zlotkin, S. H. (2003) An evidence-based approach to the development of national dietary guidelines. J. Am. Diet. Assoc. 103: S22-S27.

5. Myers, E. F., Pritchett, E. & Johnson, E. Q. (2001) Evidence guides vs protocols: what's the difference? J. Am. Diet. Assoc. 101: 1085-1\190.

6. Splett, P. (2000) Developing and Validating Evidence-Based Guides for Practice: A Tool Kit for Dietetics Professionals. The American Dietetic Association, Chicago, IL.

7. Institute of Medicine (1994) How Should the Recommended Dietary Allowances be Revised? National Academy Press, Washington, DC.

8. Institute of Medicine (2002) Protein and Amino Acids. In: Dietary Reference Intakes: Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids, pp. 10.1-10.143. The National Academies Press, Washington, DC.

9. Institute of Medicine (2002) Dietary, functional, and total fiber. In: Dietary Reference Intakes: Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids, pp. 7.1- 7.47. The National Academies Press, Washington, DC.

10. National Nutrition Monitoring and Related Research Act of 1990 (1990) Public Law 445, section 301. 101st Congress, 2nd session, October 22.

11. Clinical Practice Recommendations [Online]. American Diabetes Association. http://www.diabetes.org/for-health-professionals-and- scientists/cpr.jsp [accessed July 3, 2004].

12. Nutrition Labeling Education Act (NLEA) (1990) Public Law 101- 535.

13. Guidance for Industry: Significant Scientific Agreement in the Review of Health Claims for Conventional Foods and Dietary Supplements [online]. FDA (Food and Drug Administration) http:// www.cfsan.fda.gov/~dms/ssaguide.html [accessed July 3, 2004].

14. Consumer Health Information for Better Nutrition Initiative Task Force Final Report [online]. FDA. http://www.cfsan.fda.gov/ ~dms/nuttfloc.html [accessed July 3, 2004].

15. West, S., King, V., Carey, T. S., Lohr, K. N., McKoy, N., Sutton, S. F. & Lux, L. (2002) Systems to Rate the Strength of Scientific Evidence. Evidence Report/ Technology Assessment no. 47. AHRQ (Agency for Healthcare Research and Quality) Publication No. 02- E016. Rockville, MD.

16. Guidance for Industry and FDA: Interim Evidence-based Ranking System for Scientific Data [online]. FDA. http://www.fda.gov/bbs/ topics/NEWS/ 2003/NEW00923.html [accessed July 3, 2004].

17. Office of Federal Register (2004) Code of Federal Regulations: Food and Drugs. 21 CFR 101.14 (a)(2). U.S. Government Printing Office, Washington, DC.

18. Institute of Medicine (2002) Evolution of Evidence for Selected Nutrient and Disease Relationships. Committee on Examination of the Evolving Science for Dietary Supplements. National Academy Press, Washington, DC.

Joanne R. Lupton2

Faculty of Nutrition, Texas A&M University, College Station, TX 77843

2 To whom correspondence should be addressed. E-mail: Jlupton@tamu.edu.

Copyright American Institute of Nutrition Feb 2005

Source: Journal of Nutrition, The

More News in this Category