America's Food and Drug Administration (FDA) said last week an ongoing investigation, also known as work by independent scientist James Neal-Kababick, showed its tests for benzene in drinks over the last 15 years may be faulty.
Kababick has now announced his Flora Research Laboratories in California have developed a new procedure that would minimise the formation of benzene during testing by halving the time and temperature needed to measure benzene levels in drinks.
Current FDA tests for benzene involved heating up the beverage, a process likely to increase benzene formation in the drink during testing and, therefore, risk unrealistic results.
Recent FDA testing has found some soft drinks containing benzene, a known carcinogen, above the maximum level allowed in US drinking water. The suspected source is two common ingredients - sodium benzoate and ascorbic acid (vitamin C) - in the drinks.
The FDA and soft drinks industry have known for 15 that these two ingredients could react to form benzene in drinks, as well as that exposing such a drink to heat could significantly raise benzene levels.
Kababick said that, bearing this in mind, "it took me 10 minutes to realise the problem with their [FDA] testing method".
His new method has now been contracted by the US government's Office of Dietary Supplements to help it examine benzene formation in liquid supplements and vitamin drinks that also contain sodium benzoate.
It could also help soft drinks firms, and producers of other liquid foods or supplements containing the two ingredients, to examine their own products.
Kababick said: "A different sampling technique to remove benzene means we can measure levels down to 500 parts per trillion, which makes it more sensitive than the FDA method. It is important for producers to see benzene formation before it hits the EU or FDA limits for water."
The flaw identified in FDA testing for benzene does not, however, mean the agency and soft drinks firms are out of the woods.
It was unclear why the FDA had apparently not altered its testing method for benzene in the 15 years since it first made the link between benzene and the sodium benzoate-ascorbic acid combination, in 1990.
"They have lots of labs all over the country and there could be a lot of factors involved, but I'm not sure why the FDA did not address this matter," said Kababick.
And, he added, the new testing procedure should not be used to belittle the issue of benzene forming in drinks containing sodium benzoate and ascorbic acid.
Recent tests on drinks conducted by Kababick found a bottled lemon concentrate drink with 50 parts per billion benzene, five times above the World Health Organisation's limit for benzene in tap water.
"Actually I would grade that as fairly accurate," said Kababick, despite the result coming from what he termed a transition testing model.Other tests done by Kababick's Flora labs revealed "fairly low" levels of benzene in various different kinds of soft drinks.
But, he said, all the samples had been refrigerated prior to testing, and the effect of heat on benzene formation should not be discounted.
Industry testing on soft drinks 15 years ago is thought to have found that temperatures of 30°C and exposure to UV light for several hours were enough to more than triple benzene residues in some drinks.
Both Kababick and a food scientist who worked on those industry tests said it was essential for authorities and companies to test for benzene in soft drinks exposed to a range of different storage and transport conditions.
"Not all consumers are buying their beverages from refrigerators in grocery stores. And sometimes when the drinks are brought home they are left standing in the sun on a porch or in a hot car," said Kababick.
Data reported by America's soft drinks industry association in the 1980s showed that soft drinks could be exposed to between 32°C and 49°C in US summer months.
Kababick said his new testing procedure merely aimed to provide greater accuracy during tests on drinks from a range of different conditions. He expected to publish a journal article on the procedure, together with tests on drinks, within six months.
The stakes are high following recalls of drinks in the UK, and the launch last week of the first lawsuits against soft drinks firms over benzene in drinks.
Law firm McRoberts, Roberts & Rainer LLP joined forces with tobacco litigation veteran Tim Howard to file class action lawsuits against In Zone Brands, who make Bellywashers drinks, and Polar Beverages.
They alleged independent tests showed both companies had drinks contaminated with benzene above the limit for drinking water in the US.
The FDA has repeatedly said that none of the benzene levels it had found in drinks so far were considered a health risk for consumers.
The continuing presence of the issue 15 years after it was discovered, however, suggests a communication breakdown. One FDA scientist told BeverageDaily.com in February that soft drinks firms had pledged in 1990 to "get the word out and reformulate".
The American Beverage Association said reformulation did take place, but that some brands may not be aware of the potential for sodium benzoate and ascorbic acid to form benzene.
"15 years ago it was under control, but this is a fast-growing industry. There are a lot of new companies, a lot of new brands and things have changed," association spokesperson Kevin Keane told this website.
BeverageDaily.com broke the current story on benzene in soft drinks, after it uncovered in February that recent FDA testing had found some drinks containing benzene above America's limit tap water.