CSPI'S SUPPLEMENTARY RECOMMENDATIONS
IF HEALTH CANADA DISREGARDS OUR PRIMARY RECOMMENDATION TO DELAY IMPLEMENTING U.S.-STYLED HEALTH CLAIMS IN CANADA
I. APPROPRIATENESS OF U.S. GENERAL REQUIREMENTS
If Health Canada disregards our recommendations and decides to proceed with its proposal to permit health claims, we recommend the adoption of the following general requirements:
A. Government pre-clearance
If health claims are allowed, CSPI recommends that Health Canada adopt the requirement that all health claims be pre-approved by the government. We are strongly opposed to any system akin to the provisions of the U.S. Food and Drug Administration Modernization Act of 1997 (FDAMA), which allows companies to make health claims based on published authoritative statements from certain federal scientific bodies.
Under FDAMA, manufacturers must notify the U.S. FDA at least 120 days in advance of marketing a product with the prospective health claim, demonstrate that the claim is based on an authoritative statement, and submit a balanced representation of the scientific literature on which the claim is based. The claim can then be made unless the FDA, within the 120-day period, issues a regulation prohibiting the claim or brings a lawsuit against the company. The entire process occurs in secrete, without the benefit of public comment from the medical, consumer, or public health communities.
If Health Canada ever adopted a similar policy and did not require government pre-approval of all health claims, it would be more difficult for the government to prohibit poorly substantiated and misleading claims before they appear on labels. This would lead to consumer confusion and undercut the credibility of the entire health claim system.
B. Significant scientific agreement standard
If Health Canada proceeds, CSPI urges that all health claims be supported by "significant scientific agreement" among qualified experts and by the "totality of publicly available evidence." This evidence should include data from well-designed studies conducted in accordance with recognized scientific procedures and principles. A high scientific standard will ensure that claims are valid and will increase consumer confidence.
Health Canada should not allow health claims based on preliminary studies, even if they are accompanied by explanatory disclaimers. The recent U.S. court decision in Pearson v. Shalala held that claims which do not meet the significant scientific agreement standard may be allowed if they include a disclaimer indicating that the claim is based on preliminary scientific evidence. (4) Health Canada should not follow the principles set out in this court decision for four reasons. First, the decision was based on the First Amendment of the U.S. Constitution, which is unique to U.S. jurisprudence. Second, the case was decided by the U.S. Court of Appeals for the D.C. Circuit, and is therefore not binding across the U.S.(5) Third, the FDA has not determined how it will implement the ruling of the case. Fourth, the principles underlying the decision are unsound. Preliminary claims may later be refuted by new studies and may reduce the credibility of all health claims, thus undercutting the entire regulatory system. Even the U.S. Court of Appeals did "not rule out the possibility" that the use of disclaimers "would bewilder consumers and fail to correct for deceptiveness."(6)
C. Need for disqualifying levels
If Health Canada allowed health claims, it should establish disqualifying levels for those nutrients that have the potential, at high levels of consumption, to increase disease risk so that consumers who rely on health claims will be consuming foods that will assist them in following the recommendations of public health authorities and constructing a healthy diet. In short, health claims should not be allowed for foods with unhealthful levels of nutrients.
A health claim must provide consumers with the assurance that the food does not have properties that are potentially harmful to health. Therefore, a product should not be allowed to highlight a characteristic that may help reduce the risk of a disease, but remain silent about another characteristic that may affect the risk of the same, or another, disease or health condition. For example, a health claim on whole milk that consumption of calcium-rich foods may help reduce the risk of osteoporosis could encourage consumption of a product high in saturated fat. Diets high in saturated fat raise cholesterol levels and increase the risk of heart disease. Health claims would only be appropriate on containers of low-fat and skim, which have all the benefits of whole milk, without the unhealthful levels of saturated fat.
Thus if Health Canada proceeds, we would support the adoption of the U.S. FDA's current rules that prohibit health claims on foods that contain high levels of fat, saturated fat, cholesterol, or sodium. However, we strongly recommend including additional disqualifying nutrients: trans fatty acids and added sugars. As Health Canada and the Nutrition Labelling Advisory Committee recognized in the discussion paper on nutrition labelling, trans fat is an important factor in raising blood cholesterol levels and increasing the risk of coronary heart disease. Moreover, as documented in a petition submitted by CSPI to the U.S. FDA, added sugars squeeze healthier foods out of the diet, thereby compromising intake of other essential nutrients and raising the risk of osteoporosis, high blood pressure, cancer, and other problems that could be avoided by eating healthier diets. For this reason, added sugars should also be considered as a disqualifying nutrient.
D. The 10% rule
In the U.S., CSPI has supported the FDA requirement that disallows foods from bearing health claims unless they have at least 10% of the Daily Value for protein, fibre, vitamin A, vitamin C, calcium, or iron. However, the fortification of foods solely for the purposes of making a claim should not be allowed. This requirement is aimed at prohibiting less-than-nutritious foods, like candy or soda, from making health claims. Encouraging the consumption of foods that provide calories but few nutrients does not help consumers improve their health. For health claims to be meaningful, they should be allowed only for those foods that are significant sources of important nutrients, fibre, and other valuable constituents.(7)
E. Emphasis on total daily diet
If health claims are allowed, then they should be stated in the context of overall dietary recommendations and should emphasize a total dietary pattern not highlight a specific food. A generic claim, like "a low fat, low cholesterol diet high in fruits, vegetables, and grains that contain soluble fibre may reduce the risk of heart disease," reinforces the advice consistently given by health authorities that only overall diets not individual foods can reduce the risk of disease. In contrast, allowing health claims for individual foods misleads consumers by portraying specific products as magic bullets. Moreover, allowing health claims for individual foods would allow other less-healthful products that contain these individual ingredients to ride on the coattails of the foods that qualify to make health claims. For example, U.S. history demonstrates that the words "oat bran" alone convinced many consumers to increase their consumption of oat bran-containing foods even though many were unhealthful and some of which contain sufficient amounts of saturated fat to actually increase the risk of heart disease.(8)
To ensure that health claims emphasize the total daily diet, we recommend that Health Canada carefully consider Sweden's Food Industry's Self-Regulating Programme regarding health claims. While we disagree with the voluntary nature of this program, it is worthwhile to note that the guidance document states that health claims "must be based on the importance of the product in a balanced diet, and must be in line with official Swedish dietary recommendations." (9) The document also states that health claims should only be made for products whose normal consumption has "a substantial effect on the diet as a whole," and should "emphasize the complete picture of a nutritionally adequate diet."(10) For example, the following claim would be permitted under the Swedish program: "High levels of cholesterol in the blood increase the risk of cardiovascular disease. The intake of saturated fat contributes to an increase in the level of cholesterol in the blood. X contains a low amount of saturated fatty acids." The following claim would not be approved: "X provides excellent protection against cardiovascular disease through its low content of saturated fatty acids."(11)
F. No split claims
Health claims should not be allowed to be split between the front and back of the food label. A recent study in the U.S. examined the effects of health claims on consumer information search and processing behaviour.(12) The study found that the presence of health (and nutrient-content) claims on food packages induces consumers to truncate their information search to the front panel of packages. The findings suggest that if health claims were split between the front and back panel, consumers would be likely to focus their attention only on the front and would therefore miss potentially important information on the back.
Indeed, because all health claim information must be reported in both official languages in Canada, wherever all of the health claim information cannot fit together on one side of the package, clarity might be better accomplished by splitting the English health claim from the French health claim. Thus, rather than reporting an English and French claim on the front with qualifying information in both languages on the back, manufacturers could be required to put the entire French claim on one side and the entire English claim on the opposite side.
G. Advertising Claims
In the event that health claims are allowed, Health Canada should take all reasonable steps to ensure that the above-mentioned regulatory requirements are also applied to advertising. The failure to apply identical health claim standards to advertising claims would completely undermine the effectiveness of a labelling program.
H. Preventing Manufacturers from Escaping Regulatory Oversight
In the U.S., food companies have been able to communicate implied diet-disease messages to consumers and avoid the regulations issued pursuant to the NLEA by superficially redesignating their food products as medical foods (which are governed by other statutory provisions), by making structure/function claims, or by paying health charities for product endorsements used in advertising. We urge Health Canada to coordinate any rules permitting health claims with policies governing natural health products, structure/function claims, third party symbols, or any other similar rules which could foreseeably be used by the food industry to subvert the intent of minimum regulatory standards for health claims.
(1) Structure/Function Claims
Although structure/function claims are outside the scope of this discussion, we believe it is important to emphasize that they be subject to the same regulatory requirements that we have proposed for health claims. Since almost all structure/function claims implicitly suggest disease prevention and/or treatment, and are beyond the ability of consumers to evaluate, such claims should be regulated as strictly as health claims.
In a study on health claims commissioned by the National Institute of Nutrition, many Canadian focus group respondents tended to confuse health claims with nutrient content claims. (13) Since the difference between structure/function claims and health claims is even more subtle, it is doubly important that Health Canada ensure that structure/function claims are adequately regulated.
Careful regulation of structure/function claims is necessary to prevent the controversy and confusion that has recently arisen in the U.S. U.S. food companies may make structure/function claims without FDA pre-market approval. For example, Campbell's V8 Juice claims on its label that "Research...suggests that antioxidants in V8 juice may play an important role in slowing changes that occur with normal aging." Although the evidence linking the consumption of antioxidants to aging is preliminary, the claim does not mention a specific disease and is phrased as a structure/function claim, thereby evading FDA's rules requiring that health claims be supported by significant scientific agreement.
Manufacturers in the U.S. also make structure/function claims for products that exceed the disqualifying levels of nutrients required for health claims. For example, Kellogg's Ensemble carrot cake contains too much fat to make a heart disease claim. To get around this requirement, the company makes a structure/function claim by stating that the cake is "made with a natural soluble fibre that actively works to promote heart health." Although this claim does not specifically mention a disease, it certainly implies that the cake can help reduce the risk of heart disease. But because it technically qualifies as a structure/function claim, the U.S. FDA's rules regarding fat content do not apply.
In sum, if Health Canada decides to permit health claims, then it should establish the same strict regulatory requirements for health claims and structure/function claims.
(2) U.S. dietary supplements and medical foods as analogous regulatory categories to Canadian natural health products (currently under development).
In the U.S., an increasing number of foods with structure/function claims are also being sold as dietary supplements in order to avoid FDA approval. For example, The Hain Food Group, Inc. recently marketed a line of canned soups containing medicinal herbs, including St. John's Wort and stated: "Give your mood a natural lift." By labelling the soups as a "supplement," the company attempted to escape its obligation to obtain FDA approval of health claims.
Other U.S. food manufacturers have made similar use of the more permissive rules regarding the sale of so-called "medical foods" to avoid the legal requirements for regular foods bearing health claims. Medical foods are normally consumed under the direct supervision of a medical doctor, so the FDA did not establish specific labelling requirements for this product category. Traditional food products sold directly to the consumer, however, are now being labelled as "medical foods" in order to escape FDA regulations that apply to health claims for traditional products.
(3) Third Party Symbols
CSPI recommends that the same rigorous standards set to regulate health claims also be applied to products carrying third party symbols for health charities. Because endorsement of a food by such organizations conveys the meaning that the food has a health benefit in relation to the subject matter of the endorsing charity's mandate, the standards for determining whether a food qualifies for the logo should be at least as stringent as those set by Health Canada for other foods.
A recent National Institute of Nutrition survey found that while 46% of respondents found third party endorsements to be "very valuable," only 5% believed that such claims were very credible when they were purchased by food manufacturers.(14) Having enforceable government standards for third party symbols would eliminate some of the concerns about potential for bias in health associations setting their own minimum standards for endorsement eligibility in exchange for fees from client food companies.
II. APPROPRIATENESS OF U.S. SPECIFIC REQUIREMENTS
In the U.S., CSPI has supported the adoption of all the current U.S. health claims and their respective requirements, with the exception of the health claim concerning soluble fibre (from oatmeal, oat bran, and psyllium) and the risk of coronary heart disease. Unlike the other currently approved health claims that emphasize the effect of nutrients (e.g., sodium, saturated fat, calcium) or classes of foods (e.g., fruits and vegetables) on diet-related disease, this claim emphasizes the effect of specific foods, such as oatmeal. We oppose this health claim for the following reasons:
A. Allowing health claims for individual foods misleads consumers by portraying specific foods as "magic bullets."
Allowing health claims for specific foods such as oatmeal is inherently misleading because it portrays such foods as "magic bullets" in the fight against heart disease. Such a health claim ignores the basic principle that only overall diets not individual foods can reduce the risk of disease. As a result, consumers may be led to believe that as long as they have a bowl of oatmeal or psyllium-containing cereal for breakfast, they can indulge in cheeseburgers for lunch and still lower their risk of heart disease.
In addition, if Health Canada allows claims for foods containing oatmeal, oat bran, and psyllium, food companies may demand the right to make competing claims for dozens of other individual food items. A marketplace filled with food-specific claims will mislead consumers into thinking that the newest "magic bullet," rather than an overall healthy diet, is the key to reducing the risk of disease.
B. Allowing this health claim will lead to misleading implied health claims for other products that contain oats or psyllium as ingredients.
Permitting health claims for individual foods would allow other less-healthful products that contain these individual ingredients to ride on the coattails of the foods that qualify to make health claims. During the oat bran "craze" in the U.S. in the 1980s, sales jumped not just for oat bran cereals, but also for oat bran doughnuts, oat bran potato chips, oat bran pretzels, and even oat bran beer, all of which contained trivial amounts of oat bran and none of which were entitled to make health claims. This demonstrates that even the words "oat bran" alone would be enough of a marketing draw to convince many consumers to increase their consumption of those foods, even if they contain unhealthful levels of other nutrients.
C. The claim is unnecessary.
A food-specific claim for foods containing oats or psyllium is unnecessary since those foods already qualify to make a similar, but generically worded, claim: "A low fat, low cholesterol diet high in fruits, vegetables and grains that contain soluble fibre may reduce the risk of heart disease." However, unlike the food-specific claim, this claim reinforces the advice consistently given by health authorities that only overall diets can reduce the risk of disease.
D. The claim is misleading.
A health claim for single foods is misleading since the health benefit can only be achieved by consuming an unreasonably large amount of the food every day. For example, consumers must eat up to three servings of oatmeal every day if they rely on oats alone to improve their health. The amount and frequency with which a food must be consumed in order to attain a health benefit is particularly important in a case where a health benefit is linked to a single food. (15)
III. FORMAT ISSUES
To improve the readability of health claims in the U.S., CSPI has supported the elimination of the multifactorial statement. CSPI also supported deleting references to sex, race, ethnicity, age, and exercise in the proposed claim for osteoporosis. However, if Health Canada proceeds, we urge it to require the osteoporosis claim to state that "adequate calcium in a healthful diet throughout life may reduce the risk of osteoporosis." In addition, we recommend that independent consumer research be conducted to determine the necessity of adopting other claim requirements and the appropriateness of permitting shorter claims.
IV. CREDIBILITY ISSUES
In the U.S., the credibility of health claims is indeed a problem, as most consumers view claims on food labels as simple advertising.(16) Therefore, if health claims are allowed, we propose the addition of the following statement: "This claim is approved by Health Canada." This statement would enhance the credibility of health claims and give consumers greater confidence that health claims are based on sound scientific support.
In addition, requiring government pre-approval and establishing a high scientific standard for all health claims would ensure that the claim is valid and therefore increase consumer confidence. In contrast, allowing unapproved or preliminary claims would undercut the credibility of valid health claims and the food label itself.
4. Supra, note 2.
5. The FDA, however, may on its own accord, accede to the holding on a national basis.
6. Supra, note 2.
7. However, we believe that fruits, vegetables, whole-grain breads, whole grain cereals, and whole-grain pasta should be exempted from the 10% nutrient requirement. Diets rich in fruits, vegetables, and whole grains have been associated with a lower risk of cancer in many studies and researchers have not identified the constituents that appear to lower the risk of cancer. It is therefore reasonable to allows those foods to make health claims even though they fail to meet the 10% requirement. Multi-ingredient fruit, vegetable and whole grain products should not be exempted, however, if they contain a significant levels of non-vegetable or non-fruit ingredients.
8. See infra, pp. A-ix to A-xi for a more complete discussion of this issue.
9. Health Claims in the Labelling and Marketing of Food Products, The Food Industry's Rules (Self-Regulating Programme), (Rev. Aug. 28, 1996), at 4.
10. Id. at 7.
11. Id. at 10.
12. Brian Roe, Alan S. Levy, & Brenda M. Derby, The Impact of Health Claims on Consumer Search and Product Evaluation Outcome: Results from FDA Experimental Data, 18(1) Jnl of Public Policy and Mktg 89 (Spring 1999).
13. NIN, Health Claims in Canada: Taking the Consumer Pulse, (Ottawa: NIN, 1999) at 24.
14. Supra, note 13 at 36-37.
15. The amount and frequency are less important for claims for nutrients, because people typically get nutrients from a wide variety of foods. Similarly, the amount and frequency are less important for claims about classes of foods (e.g., fruits and vegetables) because people typically eat them several times a day.
16. 79% of respondents to the 1995 FDA Health and Diet Survey agreed (either strongly or somewhat) with the following statement: "Claims on food labels are more like advertising than anything else." Levy and Derby (1996).