Can Bird Flu Survive in Milk?

Now that avian influenza is circulating among dairy cattle in at least 12 states in the U.S. and has infected three dairy workers, health experts are keeping a close eye on whether people can be infected from consuming infected milk or meat.

So far, the federal government maintains that the risk of getting infected is low for the general public, and that commercially sold milk remains safe to drink. That’s despite the fact that U.S. Food and Drug Administration (FDA) found that about 20% of milk sold in stores contains fragments of the bird flu virus H5N1. Those fragments so far are not active, however; researchers report that they could not generate any live virus from them in the lab, and animals exposed to them did not develop infections.

Both agencies also say that pasteurization, or heating milk, inactivates the virus. But the timing of the pasteurization and the amount of virus in the milk before it's treated are important to understanding how effective heat-treating can be.

In a report published in the New England Journal of Medicine, researchers at the National Institute of Allergy and Infectious Diseases and the University of California, Los Angeles wanted to better understand how well the process can inactivate H5N1. They tested raw milk treated at two different temperatures—63°C (145°F) and 72°C (161°F)—which are typically used to pasteurize milk for retail markets.

Read More: We Are Not Safe from Bird Flu Until We Protect Farmworkers

The good news is that at the lower temperature, heat inactivated the virus in raw milk within two minutes—which means that commercial pasteurization, which generally heats milk to 63°C for 30 minutes, should be sufficient to inactivate H5N1. At the higher temperature, the virus was inactivated in most cases after just 20 seconds.

“When we did this study, there was no information on H5N1 in milk because it had never been observed before, so our starting point was building information on how well these viruses get inactivated by pasteurization,” says Vincent Munster, chief of virus ecology in the Rocky Mountain Laboratories of the National Institute of Allergy and Infectious Diseases. “This is the first study looking at the stability as well as inactivation and efficiency of heat treatment of H5N1 in the lab setting.”

While the findings are reassuring that conditions mimicking commercial pasteurization can effectively kill H5N1, the FDA and U.S. Department of Agriculture are conducting studies to verify that real-world milk treatment processes do indeed inactivate H5N1. Munster notes, for example, that the effectiveness of pasteurization is both time and dose dependent, meaning the milk needs to be treated for a specific amount of time, and that milk containing higher concentrations of virus may require longer heat exposure to kill all of the virus. Pasteurization facilities often treat milk from farms in multiple states, so batches may have varying amounts of virus. Treating them at the same temperatures for the same amount of time may not always inactivate all of the virus present, if the milk contains a high concentration of H5N1. “The next step is to confirm that industrial-scale pasteurization works the way it is supposed to work,” he says.

For now, it’s important to continue learning more about what happens to the virus as it moves from an infected dairy cow and into the milk supply. “Even with very efficient inactivation, H5N1 should not be in our milk,” says Munster. “So we should make an effort to ramp up our countermeasures to prevent H5N1-positive milk from entering dairy processing plants. If we don’t have H5N1 in the milk, we won’t have to inactivate it.”

Contact us at letters@time.com.