2004 Qualified Health Claims Research Executive Summary

Background and Research Objectives

On December 18, 2002, the Food and Drug Administration announced a new endeavor, the Consumer Health Information for Better Nutrition Initiative.  The goal of this initiative is to help Americans make sound dietary decisions by having more and better information available about conventional foods and dietary supplements, which in turn, could lead to decisions that may help improve their health and reduce risk of disease.  This initiative included agency evaluation of scientific evidence for qualified health claims describing diet or substance/disease relationships; a rating system was developed to provide a means by which the “weight of the publicly available evidence” could be assigned to one of four ranked levels (see below left).  

FDA Proposed Four Levels

Under the new initiative, the FDA also announced that the agency would develop consumer studies to identify the most effective ways to present science-based, truthful, and not misleading information to consumers.  The IFIC Foundation met with FDA in 2003 to discuss the FDA's overall consumer research program related to qualified health claims on food labels. Following these meetings, FDA felt that the IFIC Foundation could complement its work and broaden the scope of their current research objectives.  Coordinating with FDA and keeping some aspects of the IFIC Foundation research parallel to the FDA’s research, the research objectives for the IFIC Foundation Qualified Health Claims Project were as follows:

1) Measure consumer reaction to the FDA-proposed four levels of health claims on basis of: 

  •  strength of scientific evidence for claim; 
  •  overall healthfulness of product; 
  •  perception of product quality; 
  •  perception of product safety; and 
  •  purchase intent;  

 

2) Determine whether consumers differentiate between health claims (qualified or unqualified) and other vehicles for diet/health information such as dietary guidance and structure-function claims; and

3) Examine the impact of these alternative vehicles and alternative language versus qualified and unqualified health claims.

NOTE: For more information on various types of label claims, please refer to the IFIC Foundation Backgrounder on Functional Foods or the FDA, Center for Food Safety and Applied Nutrition (CFSAN), Web site on Label Claims.  

 

Exploration and Initial Findings: Qualitative Research Focus Groups 

To develop the strongest possible survey instrument for the quantitative study, the IFIC Foundation, with the assistance of Cogent Research of Cambridge, Massachusetts, conducted four focus groups in January 2004 (in Boston and Chicago) providing 8 hours of rich insights from both "general" and "health-active" consumers.  Initial findings from the focus groups were as follows:

  • Consumers had trouble distinguishing the four distinct levels of science behind the FDA-proposed four levels of health claims regardless of which of several language options were used to describe them.  In addition, when consumers viewed FDA's “report card” graphic showing all four claim levels with a check by the claim’s scientific ranking, they often connected the claim “rating or grade” with the quality of the product overall - not just the scientific evidence behind a specific diet-disease claim on the product.  While one or two people in each group tolerated the FDA “report card” graphic, most said it would deter them from purchasing a product if it contained a lower letter grade -  once again connecting the rating with the product's overall quality instead of understanding it as rating the level of scientific evidence associated with a given health claim.  Other graphic alternatives to the “report card” were offered, but none of the graphic elements tested well, including numbers, scale gradation, and star ranking.
  • Particular words such as "promising" or "inconclusive" were perceived to mean different things to different consumers.  For example, some consumers felt that "promising" implied "positive" and "hopeful," while others felt this type of language was very "flowery" and implied "marketing trickery." "Inconclusive" implied to some consumers "honesty" and "believability," while others felt the presence of this word implied extreme negativity. For others, simply having the word "may" was enough of a "qualifier" to make it appear that the claim was not the subject of Significant Scientific Agreement (SSA).
  • Structure-function claims were perhaps the most popular of all claims tested in the groups.  Most liked their brevity and general health mentions as compared to health claims that were seen as too wordy and too disease specific.  For example, several consumers mentioned that they did not like to see the word "cancer" on their food products.
  • Dietary guidance language (diet/disease messages about a whole category of foods, rather than a specific food substance) also played well, but this was less impactful for some.  For example, the statement used referred to fruits and vegetables decreasing the risk of cancer, an association of which many consumers stated they were already aware.  Still, dietary guidance language was seen as more appealing than current health claim language (whether qualified or unqualified). 

Web-based Quantitative Survey

The focus group findings proved invaluable in designing the IFIC Foundation’s Web-based quantitative survey, which was fielded in May 2004.  

The quantitative phase of the research employed a Web-based survey, using a monadic design and split-sampling techniques for various survey questions.  Using the Web-based survey, respondents could view a 2-dimensional, color product label and had the ability to view the front, side, and back panel of the product, including the Nutrition Facts panel and the ingredient list.  The products were void of any branding. 

Various claim formats or types were tested in the study.  Those formats tested by the IFIC Foundation and the FDA included: report card graphic, report card text, embedded, and point-counterpoint claims (see link below for specific examples).  Those formats tested only by the IFIC Foundation included: alternative language or claim wording, structure-function claims, and dietary guidance statements.  Three hypothetical product/nutrient/disease relationships were used as the testing contexts for the proposed four label claim levels as well as for structure-function claims: orange juice/calcium/osteoporosis; pasta sauce/lycopene/cancer; and breakfast cereal/trilinium/diabetes.  Trilinium, a fictitious nutrient, was used in order to determine the effect of four claim levels on consumer understanding of messages about a nutrient where consumer awareness was very low.  Different product/nutrient/disease relationships were tested for the dietary guidance statement exercise; these included orange juice/vitamin C/cancer; yogurt/calcium/osteoporosis; and tuna/omega-3 fatty acids/heart disease.  For a matrix of label claims and statements tested, including examples of label graphics, please click here

The sample population was 5,642 U.S. adults, 18 years and older; the sample data were weighted to match the U.S. census for gender, age, education and income.  Various statistical analyses were used including statistical means, ANOVA, and post-hoc tests.

While the Web-based quantitative survey paralleled FDA's planned quantitative research in many ways, the IFIC Foundation survey went further than the FDA research in testing whether consumers associate the "letter grade or report card" with the level of scientific evidence associated with a claim as FDA intended, or whether consumers associate the "letter grade" with product quality, healthfulness, or safety, thereby influencing purchase intent in unintended ways.  

Findings from the IFIC Foundation quantitative research on qualified health claims are as follows:
  1. Claim type and claim level, as well as perceptions of a product and/or awareness of a nutrient, collectively impact consumer perceptions. 
  2. Consumers have difficulty distinguishing among four levels of scientific evidence, especially with language-only claims. 
  3. Consumers can distinguish among four levels of science using a report card graphic -- but with negative consequences in consumer perception of product safety, quality, and healthfulness at lower level claims in some instances (report card graphic and text). 
  4. Consumers rate the scientific evidence and other attributes of a product containing an unqualified claim similar to those products containing a structure-function claim or a dietary guidance statement. 
A more detailed discussion of the findings follows.
1.   Claim type and claim level, as well as perceptions of a product and/or awareness of a nutrient, collectively impact consumer perceptions. Consumer perception of the level of scientific evidence behind a particular claim, as well as perception of various product attributes, is affected by the collective interaction of both the claim type (report card – graphic or text, embedded, and point-counterpoint) and the claim level (A-D).  In addition, consumers’ viewpoints of a particular product and/or their awareness of a nutrient are likely to influence their perceptions of product attributes as well as their likelihood to purchase a product.

2.   Consumers have difficulty distinguishing among four levels of scientific evidence, especially with language-only claims. Consumers’ inability to distinguish among four levels of scientific evidence using both point-counterpoint and embedded systems is demonstrated by using a card-sort type exercise.  This exercise was chosen to simulate consumers’ ability to rank four levels of scientific evidence in a multi-claim environment. Seventy-eight percent of consumers ranked the four levels incorrectly.   When asked if it was easy to distinguish among the four levels of claims, only 26 percent of consumers stated that it was easy. However, only 32 percent of those consumers who stated that it was easy, ranked all four label claims correctly.  Finally, “health active” consumers were no more able to rank the four label claims correctly than general consumers.

3.   Consumers can distinguish among four levels of science using a report card graphic -- but with negative consequences in consumer perception of product safety, quality, and healthfulness at lower level claims in some instances (report card graphic and text). Consumers appear to distinguish among the four levels of scientific evidence when presented with a “Report Card” Graphic (A-D) (see Slide A – green line).
    Slide A. Consumer Perception of Scientific Evidence (Report Card Graphic)

    NOTE: Slides A through J below illustrate the impact of various claim levels and types/formats on consumer perception of scientific evidence, other product attributes, such as product healthfulness, quality, and safety, as well as consumer purchase intent.  The mean effect is the measure of the difference in normalized scores.  The effect of product is neutralized in the following slides.

    When given a letter grade within the label claim language Text, consumers tend to distinguish only two broad levels of scientific evidence (A-B and C-D) (See Slide B – red line). Essentially, a two-level system appears to emerge.

    Slide B. Consumer Perceptions of Scientific Evidence (Report Card Text)

    Likewise, a much weaker two-level system appears when consumers view point-counterpoint claims (A-B and C-D) (See Slide C – yellow line).

    Slide C. Consumer Perception of Scientific Evidence (Point-Counterpoint)

    NOTE: In Slide C, the A-B level within the point-counterpoint system employs both “may” and “without may” textual versions as the “A”-level claim.

    When consumers view an “Embedded” claim (see link to examples for an illustration of this type of claim), they cannot distinguish among any of the four levels of scientific evidence (See Slide D – blue
    line).

    Slide D. Consumer Perceptions of Scientific Evidence (Embedded)


    Effects of Report Card format (Graphic and Text) on Other Product Attributes

    While consumers do appear to distinguish among the four levels of scientific evidence when presented with a “Report Card” Graphic (A-D) (see Slide A – green line), they also appear to carry their perception of the associated letter grade over to other product attributes.  In some instances, effects are seen with lower letter grades being associated with lower perceptions of product healthfulness, quality, and safety.  For example, products with Report Card Text C and Report Card Graphic D claims are perceived as being less healthful than those with a Structure-Function claim, and various B-claims (Report Card Text, Point-Counterpoint, and Alternative B2).  (See Slide E).

    Slide E. Consumer Perception of Product Healthfulness by Label Condition/Format

    In addition, Report Card Graphic and Text D claims convey lower perceptions of quality than do Structure-Function and Alternative B2 claims.  In addition, Report Card Graphic D claims also convey less quality than Report Card Text A claims.  (See Slide F).

    Slide F. Consumer Perception of Product Quality by Label Condition/Format

    Report Card Text C conveys less product safety than current Unqualified health claims (including Graphic and Text A claims), Structure-Function claims, and various B claims (Report Card Text, Point-Counterpoint, Alternative B2).  (See Slide G).

    Slide G. Consumer Perception of Product Safety by Label Condition/Format

    Effects of Report Card format (Graphic and Text) on Purchase Intent

    Consumers who see a product with a D-level claim (Report Card Text) are less likely to purchase such a product than those who saw a Structure-Function or Unqualified (without “may”), and a variety of B and C claims: Report Card Text B, Report Card Graphic B, Point-Counterpoint B, Alternative B2, Report Card Text C, Embedded C, and Alternative C.  (See Slide H).

    Slide H. Consumer Intent to Purchase by Label Condition/Format

    Also, consumers who see a structure-function statement are more likely to purchase a product than are consumers who see several “B,” “C,” and “D” claims: Embedded B, Report Card Graphic C, Point-Counterpoint C, and Report Card Text D.

    4.   Consumers rate the scientific evidence and other attributes of a product containing an unqualified claim similar to those products containing a structure-function claim or a dietary guidance statement. Consumers rate the level of scientific evidence and other attributes associated with a product containing a structure-function claim as similar to those products containing unqualified health claims (with or without the word “may”). It is important to note that consumers do not perceive any difference between claims that contain “may, and those that do not contain “may”, with respect to the scientific evidence associated with either claim (See Slide I – black line and light blue data point).


    Slide I. Consumer Perception of Scientific Evidence for Structure-Function Claims Matches Consumer Perception of Scientific Evidence of Unqualified Claims (with and without “may”)

    In addition, consumers do not perceive a difference among Unqualified health claims (with and without “may”), Structure-Function claims, and Dietary Guidance Statements with respect to the scientific evidence. (See Slide J).

     

    Slide J.  Consumers Rank Dietary Guidance Statements as High as Both Unqualified and Structure-Function Claims for Scientific Evidence

    However, consumers do rate Report Card claims (both Text and Graphic) at both the “A-” and “B”-level as having more scientific evidence than either Structure-Function or Unqualified health claims (with or without “may”); while “C”-level Report Card Graphic and Text claims convey a similar amount of scientific evidence as do Unqualified health claims.

    With regard to other product attributes, consumers rate products with Structure-Function claims as high as those with any of the Unqualified/“A” level health claims tested for healthfulness, quality, and safety.  In addition, consumers are just as likely to purchase a product with a Structure-Function claim as they are to purchase a product with any of the Unqualified/“A” level claims or Dietary Guidance Statements tested.

    Conclusions

    • Consumers had difficulty sorting out the strength of scientific evidence associated with four distinct claim levels, regardless of claim type. This may be indicative of consumer desire for simpler language on food and health, as seen in structure-function claims, dietary guidance statements, and in some instances, alternative language.
    • Emphasis on letter grades has unintended effects.  At higher letter grades, an exaggerated effect is observed with regard to the perception of scientific evidence, pushing consumer perception of scientific evidence of these claim formats (graphic and text) above current unqualified claims.  At the same time, “C”-level report card graphic and text claims convey a similar amount of scientific evidence as do unqualified health claims (with or without “may”).  At lower letter grades (graphic and text), negative misperceptions of product safety, quality and healthfulness, as well as negative impact on purchase intent are observed—letter grades appear to be viewed not just as an indicator of the level of science behind a claim, but rather are transferred to other product attributes.  This may potentially mislead consumers with regard to both perception and understanding of scientific evidence as well as overall diet choices.
    • Further research could determine: 1) ideal number of levels that could increase consumers’ ability to distinguish the scientific evidence associated with label claims and 2) terminology or language consumers would find most helpful in communicating the strength of the scientific evidence associated with diet-disease relationships and their respective label claims.

    Additional Resources

    US Food and Drug Administration (FDA) Consumer Health Information for Better Nutrition Initiative.

    IFIC Foundation Backgrounder on Functional Foods.

    US FDA, Center for Food Safety and Applied Nutrition (CFSAN), Web site on Label Claims.

    IFIC Foundation Qualified Health Claims Research. Matrix of Tested Label Claims, Statements, and Graphics of Products (pdf).

     

     

     

     

    On December 18, 2002, the Food and Drug Administration announced a new endeavor, the Consumer Health Information for Better Nutrition Initiative (http://www.fda.gov/Food/LabelingNutrition/LabelClaims/QualifiedHealthCla...).  The goal of this initiative is to help Americans make sound dietary decisions by having more and better information available about conventional foods and dietary supplements, which in turn, could lead to decisions that may help improve their health and reduce risk of disease.  This initiative included agency evaluation of scientific evidence for qualified health claims describing diet or substance/disease relationships; a rating system was developed to provide a means by which the “weight of the publicly available evidence” could be assigned to one of four ranked levels (see below right).  
    FDA Proposed Four Levels
    Under the new initiative, the FDA also announced that the agency would develop consumer studies to identify the most effective ways to present science-based, truthful, and not misleading information to consumers.  The IFIC Foundation met with FDA in 2003 to discuss the FDA's overall consumer research program related to qualified health claims on food labels. Following these meetings, FDA felt that the IFIC Foundation could complement its work and broaden the scope of their current research objectives.  Coordinating with FDA and keeping some aspects of the IFIC Foundation research parallel to the FDA’s research, the research objectives for the IFIC Foundation Qualified Health Claims Project were as follows:
    1) Measure consumer reaction to the FDA-proposed four levels of health claims on basis of: 
    strength of scientific evidence for claim; 
    overall healthfulness of product; 
    perception of product quality; 
    perception of product safety; and 
    purchase intent;  
    2) Determine whether consumers differentiate between health claims (qualified or unqualified) and other vehicles for diet/health information such as dietary guidance and structure-function claims; and
    3) Examine the impact of these alternative vehicles and alternative language versus qualified and unqualified health claims.
    NOTE: For more information on various types of label claims, please refer to the IFIC Foundation Backgrounder on Functional Foods (http://www.foodinsight.orgBackground_on_Functional_Foods) or the FDA, Center for Food Safety and Applied Nutrition (CFSAN), Web site on Label Claims (http://www.fda.gov/Food/LabelingNutrition/LabelClaims/default.htm).