A trade magazine says Consumer Reports is trying to force the FDA to strictly regulate an impurity in the manufacture of caramel colors that the agency believes isn’t harmful.
Elaine Watson reports in BeverageDaily.com (free subscription required) that Consumer Reports wants FDA to set a federal limit of 3 micrograms/day for 4-methylimidazole (“4-MeI”).
According to Consumer Reports, tests of leading soft drinks brands throughout 2013 showed a wide variation in 4-MeI levels, with some well in excess of the Prop 65 safe harbor levels, others well below, and some containing different levels in California (where Prop 65 applies) and New York (where it doesn’t).
CR toxicologist Urvashi Rangan, who directs its Food Safety and Sustainability Center, uses these data to make a policy judgment that these data do not (indeed, could not) support:
“There is no ‘safe’ level of 4-MeI, but if you have set a threshold, it should be well below the Prop 65 level (29 micrograms/day) – and more like 3 micrograms/day.*
“Safe” is a policy judgment, not something that science can determine. But Rangan correctly notes that it’s commonplace for scientists to usurp from legislators and agency heads the authority to make regulatory policy decisions:
“This is how toxicology works.”
Rangan also reveals that her policy preference is based on technical feasibility, not science:
“We know that [3 micrograms/day] is achievable as in our tests, Coca Cola has done it, whereas other brands are still way over the Prop 65 level. PepsiCo has reduced 4-MeI levels but they are still hovering around the Prop 65 limit, which in itself is too high.”
In other words, whether 4-Mel poses a human health risk at these concentrations is immaterial. As long as there is technology to achieve lower concentrations, lower concentrations ought to be required by law. That’s a defensible policy view, but it’s not defensible based on science.
Consumer Reports self-published its report on 4-Mel concentrations today, along with its demand that FDA strictly regulate. According to CR, the organization “tested 81 cans and bottles of various popular brands of soft drinks from five manufacturers between April and September 2013” purchased from stores in California and the New York metro area. In December 2013, CR tested 29 additional samples from the same areas of those brands that had initially tested above 29 micrograms/can. That is, for 29 products CR had two samples; for all others, it had just one.
This report was not peer reviewed, and rigorous peer review would have impeded publication. No inferences at all can be drawn from these sample sizes. Foregoing per review allowed CR to draw inferences anyway. For example, in the case of products purchased in the New York area, “the levels in the New York samples had come down.” It is not clear how CR refuted the null hypothesis that differences were random.
No one can credibly use two observations to show anything, but to be treated seriously advocacy groups like CR must persuade others that they are effective at moving the policy needle. Rangan follows this model closely, taking credit on behalf of CR:
“The fact that we found lower amounts of 4-MeI in our last round of tests suggests that some manufacturers may be taking steps to reduce levels, which would be a step in the right direction.”