Seven months after an outbreak of botulism linked to ByHeart infant formula, investigations are still ongoing into how the deadly contamination could possibly have made it into the product. And a new recall, this time of formula made by Nara Organics, has just been announced, with at least three babies hospitalized and treated for botulism in the U.S. These cases highlight a stark truth: Testing powdered formula for the bacteria that causes botulism is notoriously difficult.
Before ByHeart, a botulism outbreak had never been traced to commercial infant formula in the U.S. Nara said it had screened every batch of its formula for botulism, but it didn’t catch the contamination.
There’s a reason for that: spores of Clostridium botulinum, the bacterium behind botulism, survive pasteurization and persist indefinitely in dry environments, says Michael Perry, a scientist at the Wadsworth Center at the New York State Department of Health. That’s unlike Salmonella and Cronobacter—two pathogens the U.S. Food and Drugs Administration requires formula makers test for. And bacterial spores are often present in vanishingly small quantities: international experts have estimated that when contamination does occur, it may be as low as one spore per three kilograms of powder—that’s about one spore per five cans of formula.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
Complicating matters is that the spores themselves don’t produce botulism—they only make it after reaching a person’s gut, where they germinate and colonize. As a result, you can’t screen formula for the toxin. You have to find the spores themselves.
Botulism is a rare and deadly illness caused by the neurotoxin release by C. botulinum bacteria. The symptoms include muscle paralysis, which can lead to respiratory failure, and it is treated with antitoxins. The gold standard for detecting the bacteria is the mouse bioassay—an antiquated technique which dates to the early 20th century. A sample of the suspect material is injected into live mice to see if they become ill or die. It’s effective, but it requires waiting up to four days to see if they mice perish or not, and the full process can stretch into weeks.
Nara Organics used another method called sulphite-reducing clostridia (SRC) enumeration, an indicator test recommended by the International Commission on Microbiological Specifications for Foods, an organization dedicated to fighting foodborne illnesses caused by microbes. This test doesn’t detect C. botulinum directly: It screens for a broader group of related bacteria as a proxy, on the theory that elevated levels of these are a strong signal that C. botulinum might also be present. (Nara Organics declined to comment on the record for this story)
A ByHeart-commissioned study published in June by Frontiers in Microbiology concluded that SRC testing, even if it had been in place before the ByHeart outbreak, would not have prevented it. (C. botulinum testing is not currently required for infant formula—Nara Organics was testing voluntarily.) In a statement to Scientific American, ByHeart said that the company had developed a “comprehensive testing protocol intentionally built for the specificity and sensitivity needed to detect trace levels of C. botulinum found during our investigation,” called the Advanced C.Bot Safety Protocol, which it will deploy at different stages of formula production.
“We are confident that the Advanced C.Bot Safety Protocol will reliably detect spore contamination at the levels found during the investigation. No testing system alone can eliminate every risk, particularly with trace-level contaminants that may be unevenly distributed in powdered products,” ByHeart said. “The most important lesson is that food safety science must continue to evolve. We are sharing what we have learned because protecting infants requires transparency, rigorous science, and a category-wide commitment to stronger safeguards.”
Another option for testing is the PCR test, or polymerase chain reaction. This scans a sample’s DNA for the genes that encode for the botulinum toxin. It’s faster than the mouse bioassay but comes with a different trade-off: It detects the genetic blueprint for the toxin, not the toxin itself, Perry says, meaning it can flag up dead organisms as well as live threats.
A newer approach, called the endopeptidase assay, detects the toxin’s enzymatic activity. Perry says the assay “will determine if there’s active toxin present, and with that we can get results for both positive and negative samples, within roughly six to eight hours of sample receipt.”
But none of these methods are really designed to find a handful of dormant spores scattered through kilograms of powder. Perry’s lab saw this firsthand. “We were one of the first labs to actually detect Clostridium botulinum neurotoxin within a closed container of [ByHeart] infant formula,” he says. His team divided one 680-gram container into 34 samples, testing each one four to six times to make sure nothing was missed. “All that testing to only have one positive,” he says, adding that they also tested open containers that had been linked to cases of the disease, which came back negative.
Few researchers can even perform this kind of work. Because C. botulinum is classified as a select agent, meaning that the U.S. government determined it could pose a severe threat to public health if misused and so laboratories must meet strict federal requirements to handle it, which discourages many from offering testing at all. And it takes time—Perry’s team of seven was consumed with testing just tens of cans during the ByHeart investigation. A single can that tested positive, he says, took roughly five days to get preliminary results; confirming and sequencing the isolate took another three weeks.
Frank Yiannas, a former FDA Deputy Commissioner for Food Policy and Response, argues the industry cannot afford to wait for better detection science. “Given that there have been two outbreaks in seven months, we don’t have time to wait years for committees and surveys to opine on what we should do,” he says. “Companies should implement additional preventive actions now.” These might include upstream interventions, such as stricter dairy hygiene at the farm level, better screening of dried ingredients and developing powder-stage treatments that could reduce spore loads.
In 2023, the FDA urged infant formula manufacturers to consider C. botulinum when designing safety controls—but Yiannas noted that “the FDA’s investigation failed to identify a root cause” in the ByHeart outbreak. In March, representatives introduced a bipartisan bill to Congress that would require formula makers to test for C. botulinum, just as they do for Salmonella, but when a full can of formula may yield just a single positive result out of hundreds of tests, it’s not clear whether mandated testing will help solve the problem.
“Until a quantitative risk assessment is done, we should assume [botulism contamination is] a low probability, high severity event,” Yiannas says. “Just like airplane crashes are low likely, high severity events—but we expect preventive actions by airlines—so should we by powdered infant formula manufacturers.”
