Wednesday, April 30, 2014

is noncompliance with FDA definitions inherently misleading?



The FDA just issued a rule for nutritional claims about DHA and EPA. Explanation here. Some of the comments objected to the proposed rule on First Amendment grounds, and the FDA’s response is of interest. Basically, there’s a special statutory provision for nutrient claims based on an authoritative statement that identifies an appropriate nutrient level; if the FDA is notified in advance, these claims can be made unless the FDA acts to disapprove them. Other claims have to go through a different petition process. The FDA found that certain DHA/EPA claims weren’t based on the necessary authoritative statement, and thus couldn’t be used.

Among other things, the FDA stated that “[w]hen we establish by regulation particular definitions for terms (such as ‘good source’), the use of such terms without complying with the established definitions is inherently misleading, and therefore not protected by the First Amendment, because such use implies that the definitions and other statutory and regulatory requirements have been met, which they have not” (citations omitted).  Plus, proposed claims that stated or implied that there was some established daily value for DHA or EPA were false.

The FDA rejected the idea that it was required to allow the statements plus a disclaimer.  No appropriate disclaimer was identified by the commenters, “and indeed, we conclude that there is no disclaimer that could cure the fundamental flaw of the proposed DHA and EPA claims: namely, that the claims are not based on an authoritative statement that identifies a nutrient level, as required by statute.”  One comment suggested that consumer research could identify an appropriate disclaimer.  But the statutory expedited procedure wasn’t available unless an authoritative statement identifying a nutrient level has been made; it hadn’t been; as a result these claims weren’t permitted.  Consumer research wouldn’t change that, and the FDA isn’t required to conduct empirical research before barring false or misleading commercial speech.

Under the statute, a permissible claim must accurately represent the relevant authoritative statement, and “must be stated in a manner that enables the public to comprehend the information provided by the claim and to understand the relative significance of such information in the context of the total daily diet.”  The FDA determined that proposed claims based on “population-weighted” numbers wouldn’t enable the public to understand the claims’ relative significance in the context of a total daily diet.  “Population-weighted” numbers average amounts for different groups, e.g., women of childbearing age, children, etc.; the standard FDA practice instead is to use a “population coverage” number that is sufficient to cover all age and gender groups.  Using two different approaches would result in inconsistent and conflicting claims on food labels.  Such claims would be inherently misleading:

Consumers cannot make meaningful product-to-product comparisons based on such claims. The ALA claims [at issue] take place against a backdrop where all other food labeling references to nutrient levels are based on the population-coverage approach. In most situations, the reference value that results from the population-coverage approach will be higher than the reference value that results from the population-weighted approach; thus, by using the latter method, a company can in effect hold itself to a lower standard when making claims such as “good source” or “high.” 

The FDA gave examples of how this contradiction would occur just within ALA claims—one standard would allow a food to claim to be “high” in ALA while the other would allow only a “good source” claim. Not only would the presence of conflicting claims be inherently misleading, more generally, the population-weighted approach was inherently misleading in the larger context of the current labeling regime, which relies solely on the population-coverage approach.  Allowing population-weighted claims 

would create contradictory information about the meaning of “good source” when used to characterize the level of a nutrient. Even if a disclaimer or other modification were to explain that a given claim arose as a result of a certain statistical method for computing nutrient levels, this would not change the fact that terms such as “high” or “good source” would have two different meanings under this hypothetical regime. This is precisely what Congress sought to avoid when it passed the NLEA, and it is what we sought to avoid when we issued regulations under that statute, defining terms such as “high” and “good source.” 

Even if the claims weren’t inherently misleading, barring them would still be ok under Central Hudson.  Assuming counterfactually that this was truthful commercial speech, the government’s substantial interests in “promoting public health, preventing inconsistent and contradictory labeling claims (and thereby preventing consumer confusion), and maintaining the integrity of the food label so that consumers will have access to meaningful information that they can understand in the context of a total daily diet and that will enable them to make meaningful product-to-product comparisons so they can select foods that can lead to healthier diets” would justify the regulation. The FDA’s rule requiring consistent definitions and consistent labeling directly advances these government interests. And there was a reasonable fit between the government’s ends and its means: its approach was narrowly tailored to advance its interest in “preventing inconsistent and contradictory claims, maintaining the integrity of the food label, and promoting the public health.”  Claims using the population-coverage approach were not, at this time, being barred, and the commenters failed to explain why using “multiple, inconsistent statistical methods that generate inconsistent and contradictory claims would be preferable for consumers.”  Again:

No disclaimer would cure the fundamental flaw presented here: that the use of two different daily values for ALA would render the nutrient content claims that were based on those reference values inconsistent with one another, and would therefore impede consumers’ ability to make meaningful product-to- product comparisons based on those claims. A disclaimer cannot bring clarity to a situation where a fundamental contradiction remains. 

The FDA noted that producers could state the amount of a nutrient, but the purpose of nutrient content claims was to use contextualizing words such as “high” and “good source,” “which, because they are defined by regulation, place that type of information in the context of the total daily diet.  This purpose is only served if the terms … are given a consistent meaning.”  

Note: it is almost always possible to define a standardization measure as an anti-deception measure.  Commercial speech libertarians wouldn’t let that bootstrapping stand—giving the government the power to set the standard, and then claiming that deviations from the standard are confusing—but I think it’s a pretty good idea.

No comments: