than did people older than 50 (see Barry in Chapter 1). Pregnant women had the highest case fatality (the number of deaths among people with clinically diagnosed illness) of any group in this country, a phenomenon that has been reported in other influenza outbreaks (see Barry in Chapter 1). Local estimates of case fatality varied widely across the globe and in some circumstances (e.g., among populations never before challenged by influenza and troops stationed in close quarters) reportedly exceeded 20 percent.
Despite the devastation caused by the 1918 virus, it produced what was in many ways a typical influenza pandemic (see Taubenberger in Chapter 1) (Taubenberger, 2004). Most pandemics arrive in waves, albeit generally separated by years, rather than months. In the United States, with an aggregate case fatality of 2.5 percent, more than 97 percent of people with clinically reported influenza recovered from the disease; serological studies, conducted in the 1930s on people alive during the pandemic, suggest that less than 1 percent of people exposed to the virus died of flu. Prior exposure to pandemics in the mid-1850s and around 1890 apparently provided protection against the 1918 virus, resulting in relatively low mortality in people aged 35 and older. Thus the crucial uniqueness of the 1918 pandemic lay not in its virulence, but in the disproportionate number of deaths it caused among young adults, as reflected in its famously “W-shaped” pattern of mortality (Figure S-1).
Several workshop participants are studying this trend, described by presenter Jeffery Taubenberger as “the one issue that desperately needs to have a biological explanation before we can actually draw any lessons from 1918” (Taubenberger, 2004). Hypotheses under investigation include a genetic feature of the virus that targeted young adults; an intrinsic characteristic of their immune systems that produced a deadly response to viral infection; and—perhaps most likely—a deadly interaction between this virus and the young adult immune system.
Epidemiological analyses of the 1918 pandemic further highlight the dramatic shift in age-adjusted mortality as compared with subsequent years in which influenza was epidemic (Simonsen, 2004). Such studies also show that the profound impact of the 1918 flu on young adults was not limited to the second, autumnal wave of the disease, but could be detected in the initial herald wave and in influenza seasons for several years after the pandemic’s peak. Similar age shifts in mortality also marked the two subsequent influenza pandemics in 1957 and 1968, which caused far fewer deaths than the 1918 flu.