Clostridium botulinum, the bacterium that causes infant botulism, forms spores that survive pasteurization and hide in dry powder at trace levels, beyond the reach of the Salmonella and Cronobacter tests the FDA requires for infant formula.
When Nara Organics announced a recall of its baby formula after at least three infants were hospitalized with suspected infant botulism, the company pointed to a quality control program that, on paper, sounded thorough: every batch screened for contamination. The thing it did not screen for was the microbe that put those babies in the hospital.
That gap is not a Nara-specific lapse. Under U.S. food-safety rules, infant formula makers are required to test for two pathogens: Salmonella and Cronobacter. Neither causes botulism. There is no federal requirement to test for Clostridium botulinum, the spore-forming bacterium whose toxin can paralyze an infant within hours. Most manufacturers, including Nara, screen for it only on a voluntary basis, using a method that the available evidence suggests would not have caught the contamination behind the country's other recent outbreak.
That earlier outbreak, traced to ByHeart formula and investigated since late 2025, is what made the Nara recall a story instead of an isolated incident. According to Scientific American, ByHeart is the first U.S. botulism outbreak ever traced to commercial infant formula. The two events, separated by roughly seven months, are now reshaping how regulators, formula makers, and parents think about what "safe" really means for a can of powdered formula.
The structural problem starts with biology. Michael Perry, a microbiologist at the New York State Department of Health's Wadsworth Center, has spent more than a decade building one of the few public-lab pipelines in the country capable of testing for C. botulinum in food. In his telling, the difference between this pathogen and the bacteria regulators do require testing for is basic. Salmonella and Cronobacter don't form spores that survive heat treatment. C. botulinum does. Its spores survive pasteurization, settle into dry powder, and persist indefinitely, waiting for a chance to germinate inside an infant's gut, where they produce the toxin.
That timing matters. Finished product cannot be screened for toxin, because the toxin is not made until the spores reach a baby's intestines. So a manufacturer, or a regulator, has to find the spores themselves.
At typical contamination levels, finding them is brutally hard. An international expert estimate circulated through the International Commission on Microbiological Specifications for Foods, the body that writes testing guidance for the food industry, puts contamination, when it occurs, as low as roughly one spore per three kilograms of powder, or about one spore per five cans of typical size. The gold-standard test is a mouse bioassay developed in the early twentieth century. It takes up to four days for a mouse readout and weeks for a confirmatory result. A faster PCR test can run in hours, but it detects DNA from dead organisms as readily as live threats, so it tends to overcall. An endopeptidase assay detects active toxin enzymatic activity, and Perry says results come back in roughly six to eight hours from sample receipt.
C. botulinum is also a U.S. federal select agent, a designation that restricts which labs can legally handle it. Perry's seven-person team was saturated by tens of cans of ByHeart product during the investigation. For one 680-gram container that returned a positive signal, the team divided it across thirty-four samples and ran four to six tests apiece, then waited about five days for preliminary results and another three weeks of sequencing work to confirm.
Nara Organics, for its part, said it used a different screening method on every batch: sulphite-reducing clostridia, or SRC, enumeration, a proxy indicator recommended by ICMSF that flags the broader family of related bacteria rather than C. botulinum directly. A paper published in Frontiers in Microbiology in June 2026, commissioned by ByHeart, concluded that SRC-based screening would not have prevented the ByHeart outbreak. C. botulinum was detected in samples with non-detectable SRC. The study is ByHeart-funded industry research, not independent confirmation of ByHeart's broader safety claims, but its finding about the proxy test lines up with what the public outbreak data shows. Nara declined to comment on the record.
That mismatch, between what existing tests catch and what these outbreaks were actually caused by, has become the focus of a debate in Washington. In 2023, the FDA asked infant-formula manufacturers to consider C. botulinum in their safety controls, but former FDA Deputy Commissioner for Food Policy and Response Frank Yiannas notes that the ByHeart investigation has not yet identified a root cause. "Two outbreaks in seven months, we don't have time to wait years" for committee-driven fixes, Yiannas told Scientific American. His prescription is upstream: farm-level dairy hygiene, dried-ingredient screening, and spore-reduction treatments applied to powder before it is packaged.
In March 2026, Representative Rosa DeLauro introduced the bipartisan Infant Formula Safety Modernization Act, which would mandate C. botulinum testing analogous to the Salmonella rule. The bill's premise is straightforward: if voluntary screening missed two outbreaks in seven months, a federal requirement should not wait for a third. The complication is physics. At one spore per several cans, even a mandated end-product test will, by design, miss most contaminated containers.
At trace contamination levels, no mandated end-product test will catch every contaminated can. The two bacteria the FDA's current testing panel was built for are not the ones hiding in the powder. Any new rule the agency writes will have to start from that.