New study on FDA website casts doubt on qualified health claims

The Food and Drug Administration (FDA) recently posted to its web site a new study (.pdf) showing that FDA's own policy for permitting "qualified health claims" confuses consumers and fails to work as intended.

Qualified health claims are statements on the food label that are accompanied by an honest explanation of the limited scientific support for the statement. For example, the front of the label might say, "Pomegranate juice reduces your risk of heart disease*," with an asterisk. The disclaimer might say, "*This claim about pomegranate juice has not been scientifically proven, and is based on limited evidence from rats."

You may think it strange that such contradictory statements could end up on the food label, but the mixed messages reflect a genuine food policy conundrum. Until recently, FDA permitted only health claims that have "significant scientific agreement," such as the link between calcium and stronger bones. Food companies sued, however, claiming that this policy violated their freedom of speech and prevented them from sharing important health information whose scientific certainty fell slightly short of this legal standard. Court decisions forced FDA to begin permitting health claims based on somewhat weaker evidence.

The FDA website summarizes the various qualified health claims. I'd be interested to hear your comments about the health claims that do or do not appear on the FDA site. For example, although USDA strongly supports many millions of dollars in advertising of dairy weight loss claims, which are not endorsed in the Dietary Guidelines for Americans, I couldn't find such claims as having even qualified support on the FDA site.

The new FDA research study by Brenda M. Derby and Alan S. Levy, which does not reflect an official agency position, found that written qualifications expressing scientific uncertainty failed to significantly reduce consumer confidence in questionable health claims. By contrast, a "report card" style graphic may successfully communicate some distinctions in levels of scientific certainty. Unfortunately, however, a report card with a "B" or "C" rating seemed to increase consumer confidence in the health claim by comparison with no report card at all -- even though having no report card means the claim meets FDA's stronger standard of significant scientific agreement. The authors speculated that the qualified health claims report cards might work better if the graphic appeared on all products with health claims, even those that rate an "A."

The Center for Science in the Public Interest reported yesterday that FDA previously refused to make the new study public -- even to members of Congress! -- and finally posted the study to the agency website only after a Freedom of Information Act request from the consumer group. According to CSPI legal director Bruce Silverglade, "The FDA’s current policy allows companies to dupe consumers into thinking that this food or that food is the key to reducing the risk of cancer and heart disease.... The courts have made it clear that the First Amendment is not a license to practice quackery, but that is exactly what the Bush FDA policy sanctions."

The new report generally agrees with (but doesn't cite) the recent dissertation (large .pdf) by Ratapol Teratanavat, a clever student of Professor Neal Hooker at the Ohio State University. Teratanavat also finds that a report card graphic (see picture) strongly communicates differing levels of scientific certainty, while text qualifications seem to confuse the reader.



The research shows that qualified health claims work only if they include a graphic, like the "report card," which makes crystal clear to consumers that the health claim is unproven. In practice, report cards would probably never appear on real products with a "C" or "D" rating. The manufacturer would just decline to make the health claim. That seems to me a good outcome from the public interest and public health perspectives. The government is in a questionable position if it denies manufacturers their rights to speech, but it is in a strong position if it claims a bit of space on the food label to tell consumers when the manufacturer is lying.

Update 10/4/2005: Here's an interesting additional comment in July 2004 from a senior executive at the Ketchum public relations firm. It concludes,

Thus, the FDA’s support of qualified health claims will lead only to less informative, less instructional and less educational claims with less competitive value to marketers. Qualified health claims likely will add noise instead of clarity to the message stream, undermine the credibility and authority of regulatory statements and, potentially, even encourage retailers to place higher premiums on shelf space for products perceived to have more appealing claims. Here’s the bottom line: It is not “more information” but “more reliable sources of information” that impact consumer acceptance. The FDA should weigh these marketing issues carefully as it moves forward to establish a more definitive policy for allowing qualified health claims.