Hard Times for Soft Drinks
Stay up to date with the latest headlines via email.
It could be nearing high noon for the soda industry. After years of repeated battering over the issues of childhood obesity and tooth decay, sugary beverages have suffered an unprecedented backlash. The New York Times reported last week that soft drink sales are down for the first time in 20 years, and sales of bottled water, juices and energy drinks are continuing to eat into the soda market.
Into this anti-carbonated climate comes a potentially bigger bombshell that could spell disaster for the industry. Last month, the FDA quietly revealed that some soft drinks were found to contain the human carcinogen benzene in levels up to 10-20 parts per billion (ppb) -- four times the acceptable limit found in drinking water. Benzene, a chemical linked to leukemia and other forms of cancer, forms in certain beverages under certain conditions, such as exposure to heat and light.
The agency immediately downplayed the risk, saying that such small amounts did not pose a significant danger to health. "Levels like that with benzene, our only concern would be lifetime consumption," says George Pauli, associate director of science and policy in the office of food additive safety.
While scientists and doctors disagree on how hazardous benzene is to human health, the Environmental Protection Agency requires public notification and alternative water supply for drinking water contaminated with levels of 5 ppb. Even "relatively short periods" of exposure at that level can "potentially cause â€¦ temporary nervous system disorders, immune system depression [and] anemia," according to the agency. A lifetime of exposure, says the EPA, can cause "chromosome aberrations [and] cancer."
The FDA has not set an acceptable level of benzene for beverages, arguing that the public consumes soft drinks and other beverages in far lower amounts than they do drinking water -- a contention that any parent of a teenager might find laughable. Younger children may have already had a lifetime of benzene consumption.
Almost as alarming as the existence of benzene in soft drinks is that the FDA knew about the problem for more than 15 years, yet never revealed it to the public or took adequate measures to fix it. Even the latest round of tests would not have been conducted if it weren't for documents posted on the internet late last year by an industry whistleblower named Larry Alibrandi. Those papers concern an undisclosed study at Cadbury-Schweppes in 1990 called Project Denver, which found that certain soft drinks, particularly diet orange-flavored sodas, had the tendency to form benzene when exposed to heat and light.
While the industry contends the problem was corrected in the most popular sodas, no public recall was done at the time. Judging from their ingredients, dozens of products now on the shelves could potentially have the same problem, including such popular brands as Sunny Delight, flavored Diet Pepsi and Fanta Orange. (The Environmental Working Group has posted a partial list of possibly risky products.)
"The question is, how much does this problem still exist today?" says Alibrandi, who is now head of American Quality Beverages, a small New York producer of health drinks. "We have hundreds of examples from the trade, and many of them could potentially be a problem. What's especially disconcerting is the products engineered for children, where it's a potentially bigger problem for them since their body mass is very small."
In November 1990, Alibrandi was working in product development at the Connecticut labs of the British company Cadbury-Schweppes, when he says he was called into his supervisor's office one morning. "He closed the door and had a very, very concerned look on his face," recounts Alibrandi. "He said that a carcinogen was found in beverages, and they were concerned because they didn't know what the source was." That same day, Alibrandi booked a flight to Florida to test samples in a special lab capable of exposing them to extremes of heat and light.
After several trials, Cadbury-Schweppes' chemists determined that the benzene was caused by a chemical reaction between the preservative sodium benzoate and ascorbic acid (Vitamin C). The effect was found to be especially prevalent in diet sodas, and shot up to even higher levels after products were subjected to extremes of heat and light. According to the documents, Cadbury-Schweppes' Diet Crush was found to contain benzene at 25 parts per billion (ppb) -- five times the acceptable EPA limit. After exposure to 16 hours of ultraviolet light at temperatures around 30 C (86 F), that level jumped to a whopping 82 ppb. Diet Slice (made by Pepsi) contained 1 ppb before exposure, and 41.5 ppb after exposure. Diet Minute Maid (made by Coca-Cola) contained less than 0.5 ppb before exposure and 4.5 ppb afterwards, the documents say.
Despite the comparatively high levels found in these cases, however, the products tested in Project Denver were never recalled. By law, the FDA is not allowed to order a recall of a product -- but it can issue a request for a voluntary recall and, in extreme cases, can order seizure of products. On Dec. 7, 1990, representatives of soft drink manufacturers met with FDA officials to share their findings. According to a memo of that meeting, they "expressed their concern about the presence of benzene traces in their products and the potential for adverse publicity associated with this problem." The FDA ruled that the problem was not large enough to warrant a recall, "agree[ing] that low ppb level of benzene found in these products do not constitute an imminent health hazard." [sic]
That finding, however, flies in the face of other beverage scares involving benzene at the time, and may have had more to do with companies' fear of damage to their bottom lines than legitimate health concerns. In January 1990, Perrier sparkling water in the United States had been found contaminated with benzene at levels up to 22 ppb. More than 160 million bottles of water were recalled worldwide, at a loss of $263 million to the company. Perrier's reputation took a hit as well, as the company was condemned for its failure to act quickly and for continuing to advertise during the recall.
A few months later, an Australian company named Koala Springs International ordered a recall in November 1990, when a Florida health agency found benzene levels of 11 to 18 ppb in its sparkling water with fruit additive -- which was formed by the same combination of sodium benzoate and ascorbic acid as in the Project Denver tests (in fact, the Koala Springs incident precipitated the tests in the first place).
Other recalls have taken place since the Project Denver findings. In the United Kingdom in 1998, Coca Cola-Schweppes ordered a recall of Malvern sparkling water, as well as cans of Coke, Sprite, Fanta and Dr. Pepper found to contain benzene at levels up to 20 ppb due to contaminated carbon dioxide. Britvic Soft Drinks shortly followed suit, recalling more than 2 million cans of soda, including Regular and Diet Orange Tango, Lemon Tango, Pepsi and 7-Up, which had also been made with the contaminated gas. At the time, the British Soft Drink Association stated that the products were being withdrawn for "quality reasons," not because they posed a health threat, but reaffirmed a vow to recall any beverages contaminated with benzene at more than 10 ppb.
And in June 1999, Coca-Cola was forced to recall 65 million cans of Coke in Belgium and France after more than 200 people became mysteriously sick. The company's initial stonewalling on the issue caused a public relations disaster that led to a 10 percent drop in stock price and temporary bans in several countries. While the company eventually determined that the contamination was due to bad carbon dioxide and pallets contaminated by a benzene derivative, a European commission later concluded that Coca-Cola's explanation was "highly unlikely," leaving lingering questions about the source of that contamination.
Apart from the potential bad publicity, Alibrandi speculates that the Big Three soft drink makers (Coca-Cola, Pepsi and Cadbury-Schweppes) didn't publicly recall their products in 1990 because of fears that they might have to replace sodium benzoate -- an important anti-microbial preservative. Without it or its cousin potassium benzoate, he says, drink makers would be unable to cold-bottle their drinks, instead having to undertake the more costly process of heat pasteurization. "The Big Three are going to safeguard that preservative," says Alibrandi. "If they told authorities the magnitude of it, maybe the risk was to have the preservative pulled. I imagine that would create a technical nightmare for these folks."
The fix is in?
After the Project Denver tests, the industry moved quickly to minimize the problem. In less than a month, Cadbury-Schweppes changed the formula for Orange Crush, removing ascorbic acid from the drink. Later, chemists discovered that the benzene-causing reaction could be slowed by a "technical fix" -- the addition of other chemicals called "chelating agents," of which the most common is called calcium disodium EDTA. "The soft drink industry promptly took steps to address the causes of benzene formation, and the matter was resolved through improved manufacturing procedures," said American Beverage Association (ABA) spokesperson Kathleen Dezio in a statement, when the whistleblower documents were posted last year.
After the most recent revelations, ABA vice president Mike Redman, who was at the 1990 meeting with the FDA, reiterated that point in a letter to the Raleigh News & Observer : "Products that contain sodium benzoate and ascorbic acid are not inherently unsafe," he wrote. "Steps can be taken, and have been taken, in the formulation process to address reactions that may lead to benzene. You do not necessarily need to remove one of these ingredients to prevent benzene."
Spokespeople for Pepsi and Coke, which makes Fanta, referred calls to the ABA. A spokesperson for Sunny Delight, Sydney McHugh, denied that the company's products were dangerous. "We have a deliberate strategy to prevent benzene from forming in any of our products," she says, adding the company has gotten a clean bill of health from independent analysis. "If we ever find evidence of benzene in any our products, we will reformulate our products."
But recently, Alibrandi says he was shocked when he pulled trade samples of hundreds of beverages and found the same combination of sodium or potassium benzoate and ascorbic acid, including some without the "technical fix" of one of the chelating agents. "I was astounded to see the number of products that contained this combination," says Alibrandi. "If this broke 15 years ago, why wasn't this rectified across the industry? The consumers of America deserve better."
Alibrandi and his lawyer, Ross Getman, alerted the FDA to the problem last November, but the agency initially denied the need for new tests, saying that it had adequately dealt with the issue in the early 1990s. To its credit, the FDA had commissioned a study of the benzene problem shortly after the Project Denver findings. In that study, which appeared in a medical journal in 1993, FDA chemists tested 50 different types of foods and beverages, including soft drinks, and found that none had a level of more than 2 ppb.
Another study released around the same time by a chemist who consulted with the FDA isolated the process whereby sodium benzoate and ascorbic acid could form benzene. In samples made to approximate soft drinks, it found benzene was formed in levels of less than 1 ppb. Even so, the study recommended "the combination of ascorbic acid and sodium benzoate in foods and beverages should be evaluated more carefully."
Other findings in the FDA's study are more worrisome. In that study, beverages were kept refrigerated, despite the indications in the whistleblower documents that results were exacerbated by heat and light. As a postscript to the study, however, researchers prepared solutions of sodium or potassium benzoate and ascorbic acid, similar to those found in some soft drinks, and exposed them to heat and light. After 20 hours at room temperature, these solutions had formed benzene in levels of 4 ppb. After another 8 days, that shot off the charts to 266 ppb. Exposing the solutions to "strong UV light" and/or temperatures of 45 C (113 F) for 20 hours shot the levels up even further, to 300 ppb. The study concluded that the "benzene formed is associated with the interaction of these two compounds. In these cases, the removal of one of the compounds may mitigate benzene formation."
Despite these findings, Pauli defends the agency's decision not to commission further testing at the time, saying that products were unlikely to be exposed to extremes of heat and light. "With the amount of staff we have, there is no way we could test more than a small sample of products," he says. "There are more important things for our people to do." Lawyer Getman, however, argues it's not unreasonable to think that soft drinks could regularly be exposed to extreme conditions. "What are they doing in New Delhi?" he says. "Many of these countries involve vendors who don't refrigerate their products. It's sold out of a cart along with the chicken kabobs."
Getman questions industry claims that all products have been reformulated to fix the problem. Because the Big Three producers and the FDA kept the benzene problem out of the press, other smaller manufacturers may have been unaware of the need for the technical fix. In addition, some European countries don't allow such chelating agents as calcium disodium EDTA, making it unclear how the Big Three's products may have been reformulated to correct the problem in those countries.
After being rebuffed by the FDA, Alibrandi and Getman organized their own series of independent tests in November, acquiring samples from as far away as Italy and Argentina and submitting them to a lab in New York. Of the dozen beverages they tested, three were found to contain levels more than 20 ppb. They sent the results to the FDA, finally alarming the agency enough to conduct its own tests.
Two weeks ago, Pauli confirmed to reporters that a small number of beverages in their study had tested positive for elevated levels of benzene up to 10-20 ppb. Since then, however, other countries including the United Kingdom, Canada, Australia, Germany and China have followed through with their own tests. Last week, tests in Britain returned more alarming results: of 230 beverages tested, 130 had benzene levels in excess of the European Union Limit for drinking water of 1 ppb, with some containing up to eight times that limit, according to The Times of London.
Neither American nor British authorities have so far released their testing results, and the FDA has yet to make a public announcement about the danger. That's unacceptable, says Tim Kropp, a senior scientist with the Environmental Working Group, a watchdog organization that has called on the FDA to release data from its study. "Without the public knowing, there is no incentive to do anything," he says. "Industry doesn't move unless they have to."
After all, says Kropp, if the public had been notified back in 1990, the current scare might have been prevented. "We've known this is a problem for over a decade, and it hasn't been fixed. This is what happens when you have a voluntary agreement that is not even made public. It boggles my mind that anyone would think that would work."
A good start to preventing future problems, says Kropp, is to set levels for harmful chemicals like benzene for food and drink similar to those that are in place for drinking water. "Benzene doesn't care whether you are drinking soda or water, and neither does your body," he says. Lawyer Getman agrees. "Consider, which does the average 5-year-old drink more of, pop or water?" he says. "You are not going to find a parent who says my kid drinks eight glasses of water a day."
Getman and Alibrandi are now awaiting the results of further testing in the United States and other countries to determine the extent of the problem that was first discovered in a lab 16 years ago. As more details about what the industry did and didn't do emerge, there is a possibility that companies could be held legally at fault, adding another crisis to a soft drink industry that has had no shortage of bad news. Getman ticks off a long list of legal questions presented by the issue, including product liability and deceptive consumer practices. "Especially in hot climates abroad where no technical fix was put in," he says, "the potential implications for liability are huge."