March 8th, 2006

FDA Knew of Benzene Problem in Sodas Years Ago

By Suzanne Havala Hobbs
Charlotte Observer

Could soft drinks cause cancer?

It’s a question that deserves attention following the disclosure that some soft drinks contain the cancer-causing chemical benzene.

The news hasn’t gotten much attention.

But a science administrator at the U.S. Food and Drug Administration confirmed for me that recent government tests found benzene in soft drinks purchased off grocery store shelves. Long-term exposure to benzene is associated with higher rates of leukemia.

What makes this situation scandalous is that the FDA knew about the issue 14 years ago. The agency left it to industry to address the problem.

Even now, FDA says it would prefer that industry voluntarily get benzene out of our colas.

How is benzene getting in?

Benzene forms when sodium benzoate or potassium benzoate, used to inhibit the growth of bacteria, reacts with ascorbic acid, also called vitamin C.Glen Lawrence, now a biochemist and professor at Long Island University, performed the FDA tests that documented the problem in the early 1990s when he worked for the agency as a science advisor.

“People at FDA who were testing foods told me they found benzene in orange soda,” Lawrence said. “I said, `I think it must be coming from the sodium benzoate.’ “

Lawrence said he was able to demonstrate that sodium benzoate and ascorbic acid in soft drinks reacted to form benzene. He published his findings in the Journal of Agricultural and Food Chemistry in 1993.

Industry agreed at the time to spread the word among drink manufacturers and to reformulate their products, reported the online trade publication BeverageDaily.com.

Follow-up tests found no benzene in soft drinks.

But that has changed.

“An issue we thought went away in the 1990s has come back,” said George Pauli, associate director of science and policy in the FDA’s Office of Food Additive Safety.

FDA officials were recently alerted to the problem by a lawyer in New York who has campaigned to remove soft drinks from schools. BeverageDaily.com reported in February that a former industry scientist turned whistleblower helped organize tests that rediscovered the problem.

The FDA responded by going shopping, Pauli said. They went to a store, bought sodas and tested them for benzene.

“We found occasional levels significantly higher than expected,” Pauli said.

How much is too much?

The federal standard for benzene in drinking water is 5 parts per billion. The soft drink samples tested by FDA contained higher amounts, Pauli said.

There is no federal standard for benzene in soft drinks aside from the one used for water.

Pauli said that since the BeverageDaily.com report, he has received calls from industry representatives about the situation.

“It’s got the soft drink industry’s attention,” he said. “I expect they’re looking at this now.”

A spokesperson for the American Beverage Association did not return a phone call seeking comment on this column.

Pauli said he expects FDA will take action, but the agency prefers to seek voluntary action. He says voluntary agreements often can be reached faster than instituting new regulations.

Pauli makes a valid point. But this is a problem that was first discovered almost 15 years ago. It’s not good enough for our government to rely on unannounced, voluntary agreements with industry representatives to remove a cancer-causing chemical from such a widely consumed product.

Our elected representatives in Congress and in the executive branch should ask hard questions about why regulations have not been established requiring reformulation of drinks to prevent the creation of benzene.

Meanwhile, readers may want to scrutinize drink labels and avoid products that contain both sodium or potassium benzoate and ascorbic acid. According to Lawrence, citric acid is not a problem.

Lawrence says industry shouldn’t include the three ingredients in products.

“If they know this has a chance to form benzene, they should leave it out,” he said.

Comments

Add your own Comment

(optional)