The FDA knew bottled water was contaminated, but didn't inform the public.

The government’s May 2018 report on Sweet Springs Valley Water Company, a bottled water manufacturer in West Virginia, was alarming. An inspector from the Food and Drug Administration, during a review of Sweet Springs’ test records, found that several months earlier the company had bottled and distributed water from a source contaminated with E. coli, a potentially deadly bacteria.

When E. coli is detected in source water, companies must cease bottling until they produce five E. coli-free samples over a 24-hour period, according to the FDA, which regulates bottled water in the U.S. But according to the 47-page report—which details numerous other issues and was obtained by Consumer Reports through a Freedom of Information Act request—the company had not stopped production. Nor had it conducted any follow-up tests of the source water.

The inspector asked Sweet Springs’ manager, Mable Cox, whether she had considered recalling the bottled water, the report says. Cox had not. “She also stated the water had probably been consumed by this point, but she would conduct a recall” to alert people who might still have the water at home “if the FDA wanted her to.” That didn’t happen, either.

More On Bottled Water

“FDA did not request a recall,” Amanda Turney, an agency spokesperson, said in an email, without explaining why. Asked about the incident, Cox told CR that, after the inspection, she sent a driver to check stores, but there wasn’t any product left on shelves and she never notified consumers about the problem.

Most bottled water on the market appears to be safe. But the case of Sweet Springs isn’t an anomaly, according to a CR review of hundreds of pages of inspection records and interviews with regulators in seven states. Over the past decade, the FDA has cited companies at least 14 times for failing to meet federal quality standards for bottled water, but in most of those cases, the agency didn’t force a recall of products that might not yet have been consumed. For those products that weren’t recalled, bottlers also don’t appear to have informed consumers about test results showing excessive contamination—something that public water supplies have to do for the same sort of violations.

In other words, contaminated bottled water can still reach unwitting consumers, even if the FDA knows about the problem.

“Companies should not be selling contaminated bottled water, so whenever their own testing finds a violation, they should recall that water,” says Michael Hansen, Ph.D., a senior staff scientist at CR. “If they fail to do so voluntarily, the FDA should require them to.”

Self-Policing

Many consumers buy bottled water on the assumption that it’s safer than what flows out of their tap. That has helped fuel the growth of the bottled water industry, which reached $31 billion in sales in 2018. Forty percent of Americans believe bottled water is safer than tap, a recent CR nationally representative survey found, and about 1 in 6 don’t drink their home tap water.

But the bottled water industry is by and large self-policing. While the FDA requires bottled water to be free of E. coli and sets limits for numerous other contaminants, the agency generally doesn’t test the water itself. Instead, it relies on bottlers to periodically conduct their own tests and to keep those records on hand for FDA inspectors when they visit.

FDA inspections of bottled water facilities, though, declined by 33 percent between 2008 and 2018, according to data provided by the agency, and several years can pass between visits—meaning, as with Sweet Springs, the FDA might not become aware of contamination until long after it happened. Even when tests show contamination, in most cases manufacturers don’t have to stop bottling or alert the public—for instance, by issuing a press release.

Instead, before tests are reviewed by government inspectors, the FDA’s only apparent mechanism for alerting consumers is by requiring bottlers to put a statement of substandard quality on the label—saying, for example, “Contains Excessive Arsenic.” It’s illegal to sell contaminated products that lack this kind of labeling. But, perhaps for obvious reasons, the FDA could not point to a single example of a company ever complying.

The industry and the FDA say that numerous safeguards are in place to keep harmful products from being consumed.

The FDA has authority to issue mandatory recalls, but, Turney said, “the level of risk posed by these hazards” in many of the products CR identified don’t meet the required threshold. When the FDA learns that a product is contaminated, Turney said, it evaluates the information on a case-by-case basis to determine “the potential public health concern, and whether to take further action.” And, she said, companies must inform the agency within 24 hours when they detect a contaminant that will “cause serious adverse health consequences or deaths.”

Jill Culora, vice president of communications for the International Bottled Water Association, an industry group, said bottled water that does not meet federal regulations is subject to FDA enforcement actions, including recalls, warning letters, and product seizures, which help “ensure” that adulterated products don’t reach consumers. Asked about the incidents CR identified, she said, “It’s IBWA’s position that any bottled water product that doesn’t meet the FDA’s standard of quality for a regulated substance should not be allowed to be sold.”

But water quality researchers say the government’s regulation of bottled water can easily leave the public in the dark about contamination issues.

Erik Olson, senior director for health and food at the Natural Resources Defense Council, who authored a major study on bottled water quality (PDF), describes the bottled water regulatory scheme as a “paper tiger” with “extremely poor” disclosure requirements. “Even where disclosure is required,” he said, “it appears it’s rarely, if ever, done.”

Standards for Tap Water Differ

Unlike bottled water, municipal water systems must notify their customers of test results showing contamination that could pose an immediate health risk, according to the Environmental Protection Agency, which regulates tap water. And the public must be alerted, such as through the media or by mail, when contamination is found in their tap supply. For the most serious violations, like potential waterborne diseases, municipal systems must notify the public within 24 hours; with other contamination issues, they have up to 30 days.

The EPA routinely cites tap water suppliers for not meeting the requirements: More than 3,000 community water systems, servicing 2.9 percent of the U.S. population, violated the Public Notification Rule during the most recent quarter for which EPA data is available. But tap water suppliers at least have a legal requirement to quickly inform the public “whenever this is an acute health risk,” Olson said. “It appears for bottled water that type of notification is not being done. That’s really worrisome.”

One of the pitfalls of the FDA’s policing policy is that it relies on bottlers to diligently record their test results and to act quickly when contamination is found—but beyond the failure to notify customers about water quality issues, bottlers also don’t always carefully review their own test results.

Take the case of SweetSommer, another West Virginia bottler. In September 2009, an FDA inspector visited the company’s plant for a review. The inspector discovered that over a year earlier the company’s tests for already-bottled water revealed positive samples of E. coli. When asked, the executives said they were “unaware” of the results, according to a copy of the inspection report.

West Virginia’s state health department also participated in the inspection. FDA spokesperson Turney said that, while the FDA did not request a recall, the state agency ordered the seizure of all finished bottled water in the facility and asked the firm to conduct a voluntary recall. But according to the inspection report, the voluntary recall pertained only to products distributed in 2009—not the time frame when the positive E. coli sample was taken.

The FDA issued SweetSommer a warning letter in January 2010 about violations found during the inspection, including the failure to test for microbiological contaminants on a weekly basis and to keep two years’ worth of test records on file, as required. But the agency didn’t mention the positive E. coli sample. The company doesn’t appear to have notified the public about the issue, according to a review of news archives. A municipal water system operator that discovered a positive E. coli sample would have had to notify the public within 24 hours. SweetSommer went out of business several years ago, said Allison Adler, spokesperson for the West Virginia health department.

In a 2011 case, an FDA inspector cited a Colorado bottler, Deep Rock Water Company, after a review of test results showed the company had produced water with 20 parts per billion of bromate—a possible human carcinogen—which is twice the legal limit. But an official at the company tasked with reviewing the tests “had not noticed the result was too high,” according to the FDA’s inspection, even though the test said “Fail” and “exceeds FDA limits for bromate.” (Deep Rock was purchased by DS Services of America, a bottling company, in late 2011.)

The FDA’s Turney said only that the firm “appeared” to have corrected the bromate contamination, though the agency did not collect any samples to test and confirm. No recall was conducted. Turney said the water had been distributed for sale to consumers. A spokesperson for DS Services said the company can’t comment on what happened as a result of inspections that occurred before it had acquired Deep Rock.

In some cases, the FDA said it doesn’t crack down on companies because, though the test results may be above the official cutoff, they aren’t high enough to pose a serious health risk. Two years after the DS Services case, the FDA cited an Ohio bottler, Magnetic Springs Water Company, whose test results for bottled spring water showed 14 ppb of bromate, but it didn’t push for a recall, even though the FDA’s report noted that the contaminated product had been distributed to stores. Jeff Allison, president of Magnetic Springs, said the company didn’t notify consumers about the results because it was not required by the FDA, and a follow-up test completed two weeks later showed levels of bromate within the allowable limit.

Yet other bottlers have recalled water with bromate levels similar to what the records for Deep Rock and Magnetic Springs showed. In 2005, Starbucks recalled 4.1 million bottles of Ethos, the company’s water brand, after discovering that a supplier’s tests showed levels of bromate above federal limits. A year later, Wegmans recalled its private-label bottled water brand after tests revealed as much as 25 ppb of bromate. FDA records show at least eight bottled water recalls have been issued in the last 15 years for exceeding the 10 ppb bromate limit.

Little Follow-Up

Companies sometimes dispute test results showing excessive contamination, but instead of taking independent samples to settle the question, in several cases CR reviewed, the FDA appears to just take the manufacturer’s word.

In 2015, for instance, an FDA inspection of test records from Famous Ramona, a bottler in California, showed the company’s water products had between 1.1 and 1.7 ppb of phenols, a potentially dangerous chemical. The federal limit for phenols in bottled water is 1 ppb.

An executive at Famous Ramona told the FDA inspector he questioned the test’s accuracy, records show. And the inspector’s report asserted that the levels—if accurate—didn’t pose a human health hazard anyway. FDA spokesperson Turney told CR the company’s tests didn’t capture a representative sample of the water in question. But no samples were taken by the agency to form a definitive conclusion. Mark Filer, vice president at Famous Ramona, said in an email that the company didn’t inform consumers about the issue, and claims that the lab had an issue with “mishandling the samples.” Phenols have not been detected since, he said.

Similarly, in June 2018, the FDA found tests by DS Services of America suggested that the company produced bottled water products a month before with fluoride levels possibly up to 26 times the legal limit. While some fluoride in the water is considered a good thing because of its cavity-fighting properties, too much can actually damage teeth and bones.

The DS Services water with high levels of fluoride was shipped “without further evaluation,” according to FDA records. Turney said that the FDA couldn’t determine whether the high results were accurate and that there was not enough evidence to support a recall, but no samples were taken for further analysis by the FDA inspector during the inspection. The company, she added, later fired its quality assurance manager and retrained workers. Cynthia Millane, a spokesperson for DS Services, attributed the issue to a “record-keeping error” and claims that a company technician had “inadvertently recorded the wrong fluoride reading.” Regarding the quality assurance manager, she said the company can’t comment on personnel matters.

“This record-keeping error did not trigger a product recall, as this was not a public health safety issue,” she said. “We took corrective actions, including technician retraining and implementation of an electronic record-keeping system for results conducted by the technician.”

CR’s Hansen said the FDA should always take a sample as a follow-up to inspections that reveal test results showing contamination. If the agency finds contaminated water, he said, “it should require a recall of all water produced from the date of the initial violation until the problem is fixed.”

Delayed Action

Even when the FDA decides to push a company for a recall, removing products from the market can take weeks or longer.

For example, last April CR tested Peñafiel Mineral Spring Water, owned by Keurig Dr Pepper, and found levels of arsenic above the FDA limit. We immediately shared the results with the company and the FDA. Keurig paused production for two weeks while it investigated the issue. But the FDA waited several weeks before conducting its own tests, according to emails obtained through a FOIA request.

The FDA’s tests came back June 6, emails show, with arsenic levels similar to what CR found. But the agency didn’t tell Keurig that it believed the product should be recalled until June 14. Even then, it waited to inform the public, preferring to let the company speak first. “Obviously, if the firm issues press then that’s the ideal situation,” said one email, sent June 19.

A final decision to pull the product from stores wasn’t made until June 21, when Keurig announced a voluntary withdrawal in a press release. The FDA eventually classified the move as a recall, covering nearly 750,000 bottles of water. Keurig is facing an ongoing class-action lawsuit about the arsenic contamination. 

What You Can Do

Bottled water makers often don’t make it easy to determine the quality of bottled water, CR has previously reported.

While some companies voluntarily publish their test results or make them available to consumers on request, manufacturers aren’t required to make their test results public. And there is no central repository for reports. The International Bottled Water Association says if a company doesn’t provide test results voluntarily, consumers should purchase a different bottled water product. (Here are the reports CR has gathered so far in our investigation.)

“If you’re going to buy bottled water, I think it’s caveat emptor, buyer beware,” said the NRDC’s Olson. “There are a lot of bottled waters out there that may not be very high-quality.”

Consumers can also ask their water utility for a report on the quality of their tap water. If you’re concerned about tap water quality, a number of filters remove toxic substances, such as lead. (See CR’s water filter ratings and ream more about how to test your tap water.) State or local health departments might also offer free water test kits.