New study fails to place its limited data in perspective.

By Janet Raloff

Web edition: September 21, 2011

Enlarge

Tainted foods

This report, issued Sept. 21, points to BPA-laced brand-name foods that are marketed to kids.

Credit: BC Fund

The San Francisco-based Breast Cancer Fund has just released some provocative data on the presence of bisphenol A — a hormone-mimicking pollutant — in every brand-name canned food it tested.

Then again, it only tested a dozen cans. And considering there were two replicates of each type (one purchased in California, the other in Wisconsin), that means it examined only six foods for BPA, a constituent of food-grade plastics and metal-food-can liners.

Partially compensating for the new study’s small size, the Breast Cancer Fund argues, is that the items it focused on are “marketed to and consumed by children.”

Labels on three of those products have cartoon or Sesame Street figures, two others mention having the taste kids love and the last has a bunny on the label with kid-sized pasta inside. Since U.S. health agencies have identified developing children as being most at risk for any adverse effects of BPA, kids’ entrees and serving ware are precisely where we’d least like to find the contaminant.

Still, three soups and three pasta dishes hardly represent a reasonable cross-section of canned goods, even those typically fed to kids. So it would be hard to estimate from the values measured in these foods — from 34 to 148 parts per billion in the soups and from 10 to 34 ppb in the pasta products — a child’s weekly (much less annual) intake of foodborne BPA.

To get a better gauge of that, parents might want to consult findings of a study that we reported on four months ago (almost to the day). In that investigation, Food and Drug Administration chemists turned up the estrogen-mimicking BPA in 71 of 78 canned goods sampled.

If the Breast Cancer Fund study had been peer reviewed (which it wasn’t), reviewers should certainly have required a comparison of the newfound BPA values with previously reported amounts.

FDA, for instance, found that a number of foods that children often eat — among them canned tuna and vegetables — had BPA tainting of 300 to more than 700 ppb. Canned pasta and meat products, by contrast, tended to be on the low end of the range of BPA contamination that FDA chemists measured (and within the ballpark just reported by the Breast Cancer Fund).

The North American Metal Packaging Alliance — whose members make food cans — puts a rather upbeat spin on the Breast Cancer Fund’s findings, saying that the new data offer “further confirmation that only a very small amount of bisphenol A is found in food packaging, and those levels are well within the safety recommendations of government agencies.”

In a press release that the manufacturers’ group issued earlier this month, chairman John Rost observed: “People have heard that 93 percent of the U.S. population have BPA in them, but the mere presence of BPA doesn’t mean it is harmful.” He then added that human-exposure levels are almost always measured in terms of breakdown products of BPA in urine. And the very presence of those breakdown products, he said, “means the human body is metabolizing and clearing [BPA] efficiently and effectively.”

What this industry group fails to mention is that experimental data published a few months earlier challenge Rost’s assertion that ingested BPA necessarily will be rapidly and efficiently cleared from the body.

In contrast to most earlier studies, Cheryl Rosenfeld of the University of Missouri, Columbia, and her colleagues chronically fed mice diets laced with BPA, a situation designed to mimic real-world exposures. These animals got an estimated 13 milligrams per kilogram body weight over a day’s worth of rations. Levels of BPA later measured in blood (not just excreted in urine) were compared against blood values that developed after a single, larger dose (20 milligrams per kilogram body weight) was consumed all at once and independent of food. (Such a single, bolus dose is the model most researchers have, for convenience, employed in animal feeding trials.)

BPA levels in animals getting the one-time, bigger dose of the contaminant peaked within an hour (at around 21 nanograms per milliliters of blood). Within 24 hours it was virtually gone. But in animals getting the chronic exposure in food, BPA values peaked about 6 hours after a meal. And the peak BPA level in blood was roughly 90 percent as high as in the single-exposure treatment — despite the dose having been only 65 percent as high.

So the presence of food seemed to alter how quickly and effectively the body eliminated BPA. These findings appear in a paper posted online June 6 in Environmental Health Perspectives.

“What we found (and we cite this in our paper),” Rosenfeld says, “is that when animals are exposed through the diet, they actually show increased absorption of BPA” relative to when they get a single large dose.

Rosenfeld says her data show that with chronic dietary exposures, animals exhibit “the potential to bioaccumulate BPA over time.” And, she adds, there’s no reason to expect the human body would respond differently.

These findings weren’t referenced in the Breast Cancer Fund study, even though they might have bolstered its charge that low-level tainting of kids’ foods might pose risks.

So the new Fund data don’t appear to add much to what already had been published about BPA and canned goods. And the context in which its data were shared — a small, non-peer-reviewed, in-house report — also fails to inform risk-averse consumers about how they might want to alter dietary patterns. Other than, of course, to eschew canned goods. Moreover, some branded foods that the group holds up for censure had far less BPA tainting than many foods that FDA had examined.

The take-home here, as so often in science:  It remains an open question about how to make practical use of developing data. Potentially useful information on BPA exposures and risks are emerging in dribs and drabs. One can only hope that before long, a sufficient body of data will exist on which to base health policy and our grocery lists.