A team led by UA researcher Michael Worobey has found an answer to why so many young, healthy people died from influenza in a 1918 pandemic that killed 3 percent to 5 percent of the world’s population.

The flu was unusually fatal to people ages 20 to 40 years old, peaking at age 29 — an unusual impact for a disease that generally kills mostly the very young and the elderly.

Worobey and his fellow researchers found that the older population in 1918 had been exposed to an earlier, similar version of flu that provided some immunity to the emerging version.

Those 20 to 40 years old, meanwhile, had been exposed in childhood to a totally different form of the virus that provided no immunity to the 1918 strain.

“Our findings suggest that better understanding of how initial exposure shapes lifetime immunity may enhance the prediction and control of future (influenza) pandemics and seasonal epidemics,” the researchers wrote in a paper published Monday in the Proceedings of the National Academy of Sciences.

Worobey, who specializes in reconstructing the genetic trees of viruses such as influenza and AIDS, said the study adds to, but doesn’t replace, the standard view that flu is most dangerous for infants and the elderly.

It argues for an additional consideration for the types of immunity achieved in childhood by contact with similar viruses.

It could be that the most important question health officials should be asking is not how virulent a particular strain of flu is, but how incompatible it is to the immunity afforded different populations because of previous exposure, Worobey said.

In 1918, the population at peak risk, born between 1889 and 1893, had been exposed in childhood to a totally different virus, H3N8, that provided them no antibodies to the H1N1 strain.

Children are always at risk, Worobey said, because they have not had any exposure to flu and hence no antibodies that would protect against them.

The elderly with compromised health also remain susceptible to pneumonia and other subsequent infections, but in 1918 and other instances, they had much lower mortality rates.

“Maybe the first infection of childhood sets them up for a lifetime of fighting off similar viruses,” Worobey said in an interview Monday.

Maybe we should stop talking about the severity of new flu strains and talk more about “mismatched immunity,” he said.

Worobey said he expects his research will aid predictions of emerging strains of flu and the search for a universal influenza vaccine that would provide broader immunity than the seasonal ones created each year for specific strains.

“Universal vaccines might actually work,” he said.

The paper concludes that the 1918 influenza strain began as an avian virus that jumped to humans sometime around 1907.

It later jumped into pigs, re-emerging most recently as the 2009 “swine flu” pandemic.

Worobey, a professor of ecology and evolutionary biology at the University of Arizona, along with UA colleague Guan-Zhu Han and Andrew Rambaut, a professor at the Institute of Evolutionary Biology at the University of Edinburgh, used a mixture of genetic detective work and documentary research to pinpoint the evolution of the 1918 influenza strain.

He and Rambaut also devised a new type of “molecular clock” to measure the different rates of genetic evolution of the virus in its hosts.

Worobey said he was puzzled at first that the simulated evolution of various flu strains, which went through changes in horses, pigs, birds and humans, did not fit with historical accounts and epidemiological knowledge of flu outbreaks in human and animal populations.

Rambaut devised a new version of a molecular clock that assigned different speeds of change in hosts. Once that was done, everything matched up, he said.

That work, by the same set of authors, was published in February in the journal Nature.