Influenza is a respiratory infection caused by influenza A and B viruses, transmitted through respiratory droplets and contact with contaminated surfaces. CDC estimates seasonal flu causes 9 to 41 million US illnesses, 140,000 to 710,000 hospitalizations, and 12,000 to 52,000 deaths each year. Those numbers describe a serious recurring illness that most people accept as background risk. They are also two to three orders of magnitude smaller than what an actual flu pandemic could do.

The difference between seasonal flu and pandemic flu isn't just severity. It's the underlying biology: different antigenic profile, different population immunity baseline, different transmission dynamics, and different speed of clinical and public health response. The 1918 H1N1 pandemic killed 50 million people globally. The 2009 H1N1 pandemic killed several hundred thousand. The 1957, 1968, and 2009 pandemics all started with virus subtypes humans had no prior immunity to. H5N1 in 2024-2025 is the current scientific concern that the next pandemic flu may be brewing in dairy cattle and poultry rather than waterfowl.

Key Takeaways

What is the difference between seasonal flu and pandemic flu?

Seasonal flu is the annual circulation of influenza A and B viruses that humans have prior immunity to from past infections and vaccinations, causing predictable winter surges in temperate regions. Pandemic flu is the global emergence of a novel influenza A subtype with little or no population immunity, capable of efficient human-to-human transmission and causing illness rates far above seasonal baselines. The biology and the public health response are different problems.

Seasonal flu kills mostly older adults and those with chronic conditions, with a U-shaped age distribution: very young children and adults over 65 face the highest mortality risk. Pandemic flu often produces a W-shaped or even inverted-U distribution, with high mortality in young healthy adults. The 1918 pandemic and the 2009 H1N1 both killed unusually high numbers of working-age adults, in part through cytokine-storm-driven complications.

Population behavior also diverges. Seasonal flu produces no panic, no school closures, no supply-chain disruption. Pandemic flu can trigger all three within weeks of emergence depending on severity. Our disease severity scoring explained post covers how WHO and CDC assess pandemic potential through transmissibility plus clinical severity rather than either dimension alone.

How does pandemic flu emerge?

Pandemic flu emerges through antigenic shift: a major change in the viral surface proteins (hemagglutinin and neuraminidase) creating a novel subtype that humans have no immunity to. Antigenic shift typically occurs when influenza A viruses from different host species (humans, birds, pigs) co-infect a single host and reassort their genome segments, producing a hybrid virus capable of infecting humans efficiently while remaining antigenically novel.

Antigenic drift, the smaller-scale gradual mutation responsible for annual seasonal flu, is what drives the need for new flu vaccines each year. Drift produces small changes that partial pre-existing immunity still recognizes. Shift produces wholesale changes that pre-existing immunity does not. The 1918, 1957, 1968, and 2009 pandemics all started with antigenic shift events, creating H1N1, H2N2, H3N2, and pdm09 H1N1 respectively.

Pigs are the most-watched mixing vessels because their respiratory tract receptors support both human-adapted and avian-adapted influenza. Birds, especially waterfowl and increasingly poultry and now dairy cattle, are the natural reservoir. The current H5N1 worry is that the virus has now adapted to mammalian hosts (cattle, mink, sea lions) where it has more chances to acquire human-transmission mutations than it had in birds alone.

What does the historical record show?

The 1918 H1N1 "Spanish flu" killed an estimated 50 million people globally with a case-fatality rate of 2-3%, hitting young adults disproportionately. The 1957 H2N2 "Asian flu" killed roughly 1.1 million globally. The 1968 H3N2 "Hong Kong flu" killed roughly 1 million. The 2009 H1N1 pandemic killed around 280,000 in its first year per CDC, with relatively low CFR but high attack rate especially in young people.

Each pandemic differed in profile. The 1918 strain caused unusual lung pathology with cytokine storm and bacterial superinfection. The 2009 H1N1 was less severe per case but spread fast and infected children at very high rates, partly because older adults had cross-protective immunity from prior H1N1 exposure. Severity, transmissibility, and age distribution have all varied substantially across the four pandemics in the past century.

The interpandemic interval averages roughly 30 years, but with high variance. The flu pandemics of the past century occurred in 1918, 1957, 1968, and 2009. The next one is overdue by some metrics and not by others. Our CFR vs IFR post covers how to read mortality numbers across pandemics that differ in surveillance quality and definitional consistency.

Why are scientists watching H5N1 specifically?

H5N1 highly pathogenic avian influenza has caused devastating poultry outbreaks since 1996 and human spillover cases since 1997, with an overall human case-fatality rate above 50% in WHO-confirmed cases (a figure that almost certainly reflects surveillance bias toward severely-ill hospitalized patients rather than total infections). CDC and USDA confirmed H5N1 in over 800 US dairy herds and dozens of human cases by early 2026, all linked to occupational exposure to infected animals. The mammalian adaptation since 2022 (in mink farms, sea lions, dairy cattle) is the watch signal.

Each human case is currently a dead-end infection, with no documented human-to-human transmission. The risk is not the current case count but the genomic opportunity. Every human or mammalian infection gives the virus another chance to acquire mutations that improve human-receptor binding, polymerase efficiency in mammalian cells, and respiratory droplet transmission. The 1918, 1957, and 1968 pandemics all involved birds-to-mammals-to-humans pathways.

If H5N1 acquired efficient human-to-human transmission, the early epidemiology would look very different from current sporadic occupational cases: clusters of cases in households, healthcare-associated transmission, and detection in the wastewater of cities without confirmed animal outbreaks. Our influenza pandemic H5N1 risk post covers the technical signals more deeply.

How would a flu pandemic in 2026 be different from 1918?

A 2026 flu pandemic would have access to neuraminidase inhibitor antivirals (oseltamivir, baloxavir), better critical care including mechanical ventilation and ECMO, and pandemic vaccine production within 6 to 9 months of strain identification using mRNA, cell-based, or recombinant platforms in addition to traditional egg-based manufacturing. None of those existed in 1918. A 2026 pandemic of comparable severity to 1918 would still kill millions but at a substantially lower rate.

The headwinds run the other way too. The 2026 world is more interconnected than 1918 by orders of magnitude. International travel volume in 1918 was a small fraction of pre-pandemic 2019 levels. A pandemic strain emerging in any populated region would spread globally within weeks regardless of public health response. Modern global supply chains for medications, food, and energy are also more vulnerable to large-scale workforce illness.

Population health composition matters too. In 1918, fewer people lived to ages where chronic conditions raised influenza mortality. In 2026, the population includes far more older adults, more people with diabetes, obesity, and immunosuppression, all of which increase severe flu risk. The 2009 H1N1 pandemic saw worse outcomes in obese people than prior pandemics, partly reflecting changes in baseline population health. Our vaccine supply chain failures post covers what manufacturing capacity actually looks like.

FAQ

How does the seasonal flu vaccine differ from a pandemic flu vaccine?

The seasonal flu vaccine is reformulated each year to match circulating H1N1, H3N2, and B strains, taking 6 to 9 months from strain selection to deployment. A pandemic vaccine targets a novel strain that emerges suddenly. Manufacturing typically can't begin until the pandemic strain is sequenced and grown, meaning a pandemic vaccine would not be widely available until 4 to 8 months after the pandemic begins.

Is bird flu the same as H5N1?

H5N1 is one specific subtype of avian influenza A. Other bird flu subtypes (H7N9, H9N2, H5N6) have also caused human cases. "Bird flu" is colloquial shorthand for any avian-origin influenza causing human concern. The 2024-2025 US dairy and poultry outbreaks are specifically H5N1, with a particular genotype (B3.13) that adapted to dairy cattle, not the same H5N1 lineages causing concurrent global outbreaks elsewhere.

Can you get the flu twice in one season?

Yes, especially with different subtypes. A person can have H1N1 illness early in the season and H3N2 or influenza B illness later. Cross-protection between subtypes is partial and incomplete. The 2024-2025 US flu season had unusually high concurrent circulation of multiple influenza A subtypes plus influenza B, which increased the rate of within-season reinfections.

How effective is the annual flu vaccine?

CDC's annual estimates of flu vaccine effectiveness against medically attended illness ranged from 19% to 60% over the past decade. Effectiveness varies year to year based on strain match between vaccine and circulating viruses, particularly for H3N2 which mutates fastest. Even in low-match years, the vaccine reduces severity, hospitalization, and death in vaccinated people who do get infected, which is the relevant outcome for older adults.

Is COVID-19 a pandemic like flu pandemics?

COVID-19 met the WHO definition of a pandemic in March 2020 and continues to circulate globally, but is caused by SARS-CoV-2 (a coronavirus), not influenza. The clinical course, transmission dynamics, and treatment options differ substantially. Comparisons between COVID-19 and historical flu pandemics are useful for thinking about scale but should not be over-extended on biology. Our novel pathogens explained post covers the broader category.