Dropping science: Journalists can check food safety claims in the lab

Food is packaged with a veneer of sincerity. Contents are dutifully itemized along with tables showing the percentage of recommended nutrients, fat content, etc. But there is much that remains a mystery. You are never going to see a candy wrapper that says, "May contain lead."

I spoke with Mark Katches, the deputy managing editor for projects at the Milwaukee Journal Sentinel last week about his investigative team's testing of products containing chemicals such as bisphenol A. They produced a great story in November about cook-and-eat products, infant formula and food containers labeled "microwave safe."

The story begins:

Products marketed for infants or billed as "microwave safe" release toxic doses of the chemical bisphenol A when heated, an analysis by the Journal Sentinel has found.

The newspaper had the containers of 10 items tested in a lab - products that were heated in a microwave or conventional oven. Bisphenol A, or BPA, was found to be leaching from all of them.

The amounts detected were at levels that scientists have found cause neurological and developmental damage in laboratory animals. The problems include genital defects, behavioral changes and abnormal development of mammary glands. The changes to the mammary glands were identical to those observed in women at higher risk for breast cancer.

There are a host of promises made on food labels, and news organizations should not leave it up to Consumer Reports to truth-check them. Here are a few basic steps to consider when doing your own testing.

1. Find a lab that has experience. Don't take it on faith that they know how to check for mercury, lead, bacteria or whatever the case may be. Ask them how often they have done it, and how exactly they do it. There are often multiple ways to detect contaminants, and you want to use the method that is the most widely accepted. Often it is best to find out what the FDA uses for its testing.

2. Ask experts to vet the testing protocol. Once you know what your prospective lab is going to do, ask experts in the field about the lab's methodology. If they disagree, you can either ask the lab to make some changes or find a new lab.

3. Start small. You might start with just five samples and see if you get any hits. When we tested candy for lead at the Orange County Register, we found lead right out of the gate and decided to do more testing. When we tested hamburger for fat content, we found that the label was usually correct and decided to stop testing.

4. Pick a sample size that will withstand criticism. There is no perfect number. Obviously a bigger sample size can give you greater confidence in your results, but you might not have a budget for much more than a dozen tests. Talk with the experts who vetted your protocol about how to target your testing in the most effective way.

5. Be transparent with your audience. Post on the web a detailed explanation of how exactly you did your tests, the name of the lab, the breakdown of the results and any other important details. The Journal Sentinel posted PDFs of its microwave product test results alongside the story.