When highly pathogenic H5N1 bird flu spilled into U.S. dairy herds in 2024, veterinarians saw something strange: cows were not dying of pneumonia, but developing severe mastitis as the virus attacked their udders. At the same time, federal testing found infectious virus in raw milk from affected herds, and two human infections were tied to the dairy outbreak, raising urgent questions about how this pathogen had adapted to cattle. Scientists have now traced that unusual pattern to the biology of the mammary gland itself, and their findings are reshaping how regulators monitor milk and protect workers.
Why Bird flu attacks cows udders instead of matters now
The central shift is that H5N1 is treating the cow’s udder as its main target tissue, not the lungs. In a primary pathology study of the U.S. dairy outbreak, researchers reported that highly pathogenic avian influenza H5N1 showed a strong preference for mammary gland tissue, with viral-induced mastitis concentrated in the milk-producing parts of the udder, according to Nature. That pattern helps explain why milk, not nasal secretions, has become the key sample for tracking infection in herds.
Mechanistic work on receptors is filling in the “why.” A team publishing in the Journal of Dairy Science mapped the sugars that influenza viruses use to enter cells and found that bovine mammary tissue across physiological stages displays abundant α2,3-linked sialic acids, according to Primary. Those α2,3-linked sialic acids match the binding preference of avian influenza strains, giving H5N1 a ready foothold in the udder’s milk-secreting epithelium.
The working hypothesis emerging from these findings is that decades of breeding for high milk yield in modern dairy cows, such as Holsteins, may have intensified receptor expression in the mammary alveoli. If selection for production has increased α2,3-linked sialic acids specifically in this tissue, it would create a stable replication niche that respiratory tissues lack. Researchers could test this by using CRISPR tools to knock down these receptors in primary bovine mammary organoids and then challenging them with H5N1 to see whether replication falls compared with unedited cells.
That tissue preference is already driving policy choices. The U.S. Department of Agriculture built a National Milk Testing Strategy that treats milk as the most informative sample type for detecting H5N1 in dairy herds, according to the National Milk Testing Strategy. Because the virus concentrates in udders and milk, regulators have focused surveillance on bulk tanks and processing streams rather than relying on respiratory swabs.
The evidence behind Bird flu attacks cows udders instead of
Several independent lines of primary data now connect receptor biology, udder pathology, and milk-borne virus. The receptor-mapping study in Journal of Dairy Science reported that bovine mammary tissue displays influenza receptor-linked sialic acids, including abundant α2,3-linked sialic acids that avian viruses prefer, across different physiological states, according to Primary. That finding shows the udder is not a passive conduit for virus shed elsewhere, but a tissue inherently permissive to avian influenza entry.
Pathology work on outbreak animals goes further, showing where the virus actually replicates. In a peer-reviewed investigation of the dairy spillover, scientists documented that HPAI H5N1 in U.S. dairy cattle has strong tropism for mammary gland tissue and that viral replication occurs in milk-secreting mammary epithelial cells lining the alveoli, according to Nature. That cellular localization matches the receptor map and ties the udder infections directly to the milk that leaves the farm.
Controlled infection experiments have tested how unique this udder targeting is compared with respiratory exposure. In an experimental study of clade 2.3.4.4b H5N1, calves given oronasal doses showed moderate nasal replication and shedding but no severe clinical signs, while lactating cows that received intramammary inoculation developed severe acute mammary infection with necrotizing mastitis and high viral loads in milk, according to Primary. The contrast between mild upper-airway infection and destructive udder disease reinforces that the mammary gland is the main engine of viral amplification in cattle.
Milk testing has confirmed that this intramammary replication translates into infectious virus in the product itself. A laboratory study of raw milk from an affected herd found that samples contained infectious H5N1 isolates and that those isolates could infect mice under experimental conditions, according to Primary. Separate heat-inactivation experiments showed that H5N1 virus present in raw milk can be inactivated at 63°C and 72°C under defined conditions, according to Primary, supporting long-standing pasteurization standards while highlighting risk from unpasteurized products.
Public health agencies have linked this dairy biology to human infection. A national epidemiologic report documented an outbreak of highly pathogenic avian influenza A(H5N1) viruses in U.S. dairy cattle and detection of two human cases connected to that outbreak, with exposures that included contact with infected cattle and potentially contaminated materials such as raw milk, according to Primary CDC. Earlier, Texas health authorities reported the first case of novel influenza A(H5N1) in Texas in March 2024, describing a patient with conjunctivitis and direct dairy cattle exposure, according to the Texas health alert.
Virologists have since characterized the human virus from that outbreak. A study of the human isolate A/Texas/37/2024 found that this cattle-linked strain is transmissible and lethal in animal models, according to Texas. Those data do not show that the virus spreads easily between people, but they confirm that the version coming out of infected udders and milk has serious pathogenic potential in mammals.
Regulators have tried to translate these findings into consumer guidance. An update on the investigation of avian influenza A(H5N1) virus in dairy cattle reported that retail milk testing has detected viral fragments by PCR and that food-safety messaging centers on the expectation that pasteurization would inactivate virus in milk, according to the FDA. That stance aligns with the laboratory heat-inactivation work, while still leaving open questions about raw milk exposures for people working or living on farms.
What remains unresolved for Bird flu attacks cows udders instead of
Despite the rapid progress, key pieces of the udder story are still missing. Quantitative receptor-density comparisons between mammary and respiratory tissues from the same naturally infected cows are not yet available in the published datasets, so scientists cannot say exactly how much more α2,3-linked sialic acid the udder presents compared with the nose or trachea, based on the current Primary receptor map. Without that side-by-side measurement, the field is still working from strong associations rather than a complete hierarchy of tissue susceptibility.
Time-series data on how infection affects milk yield and how quickly mastitis resolves under field conditions also remain sparse. Reports from Texas herds describe clinical observations and sequencing context for avian influenza A(H5N1) virus among dairy cattle, according to Primary, but detailed production records across the full course of infection have not been published. That gap makes it hard for farmers and insurers to quantify long-term economic impact or plan for recovery after an outbreak is detected.
Another open question is how much breed and management influence udder susceptibility. Comparative sialic-acid profiling across dairy and beef breeds under identical conditions has not been reported in the primary literature cited here, so any link between high-yield Holstein genetics and increased α2,3-linked sialic acids remains hypothetical. Experiments using CRISPR-mediated receptor knockdown in bovine mammary organoids, followed by H5N1 challenge, would give a more direct test of whether those receptors alone explain the virus’s fixation on the udder.
Virology studies are also probing how virus in naturally infected milk behaves compared with artificially spiked samples. Work on the stability of avian influenza A(H5N1) virus in milk from infected cows and virus-spiked milk has shown that stability and heat sensitivity can differ between those two contexts, according to Primary. That suggests that conclusions about risk based solely on spiking experiments might miss features of virus that has replicated inside mammary epithelial cells.
For readers, the immediate consequence is twofold. First, the strongest available evidence indicates that pasteurized milk remains a low concern from a virology standpoint, because H5N1 in raw milk is inactivated at standard processing temperatures, according to Primary. Second, the udder-focused biology means that workers who handle live cows, milking equipment, and raw milk are closer to the main reservoir of virus than people buying cartons at a grocery store. Public health agencies are likely to keep updating guidance for dairy employees and veterinarians as more data emerge from ongoing investigations of avian influenza A(H5N1) virus in dairy cattle, according to the Primary CDC. The next phase of research will show whether changes in farm practices, breeding, or vaccination can shift the virus away from cows’ udders and reduce the risk that milk becomes a conduit for H5N1.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
