Forbes contributors publish independent expert analyses and insights.
The 1918 influenza pandemic remains the deadliest in modern history, killing tens of millions -- and leaving scientists with enduring questions about where it began.
For decades, many researchers assumed the virus emerged in pigs. But when scientists sequenced the virus's genome in 2005, they found it was primarily avian in origin. One possibility was that the virus originated in birds, passed through pigs -- where it adapted to mammalian cells -- and then jumped to humans. However, more recent evidence has cast doubt on this so-called "mixing vessel" hypothesis. Studies show that avian-to-swine transmission is rare, that pigs and humans have similar distributions of sialic acid receptors in their respiratory tracts to which influenza viruses bind, and that the 1918 virus's genome shows little sign of prolonged adaptation in pigs.
Based on this, scientists Martha Nelson and Michael Worobey concluded that seven avian genes in the 1918 virus were probably transmitted from birds to humans at the start of the pandemic, and from humans to swine once the pandemic was widespread in humans.
Now a new hypothesis, put forth by medical historian Martin Furmanski and virologist Pablo Murcia published in the Journal of Infectious Diseases, suggests a different intermediate host was spreading a prepandemic avian-origin virus before 1918: the horse.
The idea is biologically plausible and historically grounded -- and it reframes how we think about the origin of the 1918 pandemic.
Influenza viruses infect hosts by binding to sialic acid receptors in the respiratory tract. Avian viruses preferentially bind to α2,3-linked receptors, while human viruses target α2,6-linked receptors. Most mammals express one type or the other, but some -- including pigs and horses -- express both. This dual susceptibility has led scientists to propose that these species may act as "bridge hosts," enabling avian and human viruses to co-infect the same cells and exchange gene segments through reassortment. They may also provide an environment in which avian viruses can acquire additional mutations that affect processes besides receptor binding enhancing replication, cell entry, and transmission in mammals.
While pigs have long been considered for this role (although, as noted above, that notion was challenged by Nelson and Worobey), horses have only recently received attention. But a growing body of evidence shows that horses exhibit both α2,3 and α2,6 sialic acids. Moreover, all three known equine influenza viruses are believed to have derived from avian strains.
The new hypothesis draws on more than just receptor biology. It also points to an overlooked historical episode: a large epizootic of equine influenza that began in North America in 1915. Reports at the time described unusually severe respiratory illness and pneumonia among horses -- a feature also seen in human cases during the 1918 pandemic.
The timing is notable. Between 1914 and 1918, World War I drove one of the largest mobilizations of horses in U.S. history. The U.S. Army purchased over 300,000 horses and mules after joining the war effort, in addition to 68,000 used during the 1916 Mexican Punitive Expedition. Furmanski and Murcia provide evidence that over two million animals were mobilized by the military between 1914 and 1918, the majority of which were eventually exported to Europe.
These animals were housed in crowded stables, transported in train cars across the country, and shipped overseas. Such conditions created ample opportunity for viruses to spread, mutate, and reassort. Ultimately, at some point seven segments from an avian H7N1 virus combined with an H1 hemagglutinin gene that was already adapted to humans, producing the 1918 pandemic virus.
Reconstructions of 1918 influenza viruses support the idea of gradual adaptation to humans. As reviewed in a forthcoming paper in the journal Microbiology and Molecular Biology Reviews, an early strain isolated in New York has been found to bind both avian-type (α2,3-linked) and human-type (α2,6-linked) sialic acid receptors and failed to transmit efficiently in ferrets -- a standard model for human infection. Such a virus might have been well adapted to horses. By contrast, a later strain from South Carolina primarily binds to human-type receptors and transmits readily via aerosols, suggesting that it had by then acquired adaptations permitting human-to-human spread.
Camp Funston in Kansas is often cited as the site of the first well-documented outbreak of 1918 flu in humans. It was also a major staging ground for military horses. Tetanus was common, and millions of doses of antitoxin serum were produced using horses -- another point of close contact between humans and equids during wartime.
Conclusive evidence is still lacking; confirming past infections or reassortment events would require archived horse tissue or serum from the 1910s. But taken together, the historical and biological clues offer a compelling case.
Understanding how pandemic viruses emerge is key to anticipating future outbreaks. If horses can serve as a bridge between avian and human influenza viruses, they warrant closer surveillance -- particularly where they are kept near poultry or people. The recent spread of H5N1 among dairy cattle, another unexpected mammalian host, underscores how little we may yet understand about the pathways influenza can take.
Influenza pandemics are shaped by more than viral mutation. They emerge at the intersection of biology, human behavior, and historical circumstance. In 1918, that intersection may well have been in a horse stable.