How H1N1 Is Driving an Early Flu Season, and When to Get Your Shot

Last winter belonged to H3N2. Of the 48,052 influenza A viruses that U.S. public health labs subtyped over the 2025-26 season, 40,018 (83.3%) were H3N2, the family whose "subclade K" drove that season. This autumn the ratio has flipped: in the week ending Saturday, Oct. 3, 125 of the 127 influenza A viruses subtyped were H1N1.

That virus is behind a flu season running weeks ahead of the calendar. In CDC's latest weekly flu report, published Friday, Oct. 9, 3.6% of respiratory specimens tested by clinical labs were positive for flu (1,435 of 39,317), the highest reading for that week in ten seasons of CDC data. Canada reached 4.7%. The dominant strain is an H1N1 subclade called D.3.1.1, and the H1N1 piece of this year's vaccine, A/Missouri/11/2025, is a close relative. That makes waiting until winter so the shot "lasts" the riskier bet for nearly everyone right now.

Ten seasons of early-October numbers

In week 39 of 2025, clinical labs found flu in 353 of 54,629 specimens, or 0.65%. This year's 3.6% is more than five times that. Even 2022, the last season that started unusually early, sat at 2.0% in the same week.

Early October flu positivity is the highest in ten seasons

Two caveats apply. The raw count is less dramatic than the rate: this year's 1,435 positives are close to 2022's 1,351, and the percentage is higher partly because labs tested fewer specimens (39,317 against 66,897). And this is a first report. CDC's week-38 figure was first published at 3.43% on 33,547 specimens, then revised to 3.10% a week later on 45,923, as more negative tests arrived. Week 39 could too. CDC's own phrasing fits: "higher than typically observed at this time of year but remains within expected seasonal variation."

Outpatient visits for flu-like illness were 1.8% of the total, below the national baseline of 3.1%. Activity is highest on the West Coast, where HHS Region 9 hit 9.5% positivity; in Washington state the share of emergency visits diagnosed as flu rose from 0.25% in late August to 1.34% in the week ending Oct. 3, according to CDC data cited by Live Science.

Local numbers rhyme. In Louisiana, about 6.5% of flu tests came back positive that week, against 3.4% a year earlier, KPEL reported. Colorado had reported 46 flu hospitalizations as of last week, compared with 8 at the beginning of September, Denver7 reported; in that state, flu emergency visits last year didn't reach current levels until the first week of November. Maine health officials have counted 104 confirmed cases through early October, WGME reported.

Canada's picture is starker. In the week ending Oct. 3, 1,149 of 24,619 tests were positive, or 4.7%, according to Public Health Agency of Canada data. The same week in each of the past three seasons came in at 0.57% or lower, so this year is roughly eight times higher, and above even the pre-pandemic maximum for that week, 2.95%. British Columbia stands at 12.2% (448 of 3,672 tests), with outbreaks in long-term care homes driving many cases, Brian Ward of the Research Institute of the McGill University Health Centre told CBC News.

Sewage tells two stories. CDC's wastewater data rated flu "very low" as of Oct. 3, but that system is in a brief gap in updates tied to a new testing contract, USA Today reported. The Stanford-led WastewaterSCAN rated flu "high," with influenza A in 60% of samples this month against 20% in October 2025.

A different virus, and a shot built around it

The virus doing the work is A(H1N1)pdm09, the lineage many people still call swine flu, now a routine seasonal strain. Its branches, called subclades, are groups of viruses sharing particular mutations. Since week 21, 80.5% of influenza A viruses reported in the U.S. have been H1N1, CDC says, "and among these, the majority have been D.3.1.1 HA subclade." The World Health Organization said on Sept. 25 that D.3.1.1 viruses have "predominated worldwide" since February. In Canada, about 98% of subtyped influenza A since late August has been H1N1 (2,176 H1 viruses against 46 H3).

That shift matters because vaccines are aimed at specific strains. After subclade K, the FDA's advisers went further than a routine update, replacing all three components of the U.S. vaccine for 2026-27 at their March 12 meeting (the reasoning is in why every strain in this year's flu shot was swapped). The new H1N1 component is A/Missouri/11/2025.

The early evidence on fit comes from lab tests of antibodies, raised in ferrets or drawn from vaccinated people, against circulating viruses. In February, WHO reported that ferret antisera against D.3.1 viruses such as Missouri recognized the main circulating clades "well." WHO's September vaccine-strain report was mostly reassuring, with one flag. Missouri-like antisera "recognized most D.3.1 and D.3.1.1 viruses well," but showed "reductions in recognition of some viruses from subclade D.3.1.1 with additional substitutions including G155E." Those G155E viruses "increased in recent weeks," and antibody levels in vaccinated people's serum were "significantly reduced for some recently circulating" H1N1 viruses. WHO still kept Missouri for next year's Southern Hemisphere vaccine.

Britain's Health Security Agency said on Oct. 8 that a WHO technical report indicates one H1N1 virus seen at low levels in Europe "has 'drifted'" from the vaccine component, and that it is "still too early in the season to say with any certainty" whether it will spread widely.

Ward called the match good. Live Science judged it too early to know, and Dr. Paul Offit, director of the Vaccine Education Center at Children's Hospital of Philadelphia, told the outlet that when circulating strains differ from the vaccine, that usually predicts a more difficult season. Lab results are signals; real-world effectiveness is measured later in the season. For scale, CDC estimates put flu shots at 29% to 56% effective against flu illness needing medical care over the past decade.

The Southern Hemisphere, often read as a preview, is nearly silent. A PAHO study of hospitalized patients in six Latin American countries tested vaccines carrying the same Missouri component, but that season was H3N2: 3,215 H3N2 cases against just 148 H1N1. The H1N1 effectiveness estimate, 20.2%, came with a 95% interval from −32.6% to 51.9%, too wide to separate a useful vaccine from a useless one. CDC itself warns that the southern season "does not always predict what will occur in the Northern Hemisphere."

Early is not the same as severe

CDC expects the season "will likely be classified as moderate across all ages, with low to moderate confidence," and says "there is no evidence to indicate that a severe season is more likely." James Hay, an epidemiologist at the University of Oxford, has said an early start does not necessarily mean a more severe season, KPEL reported, and Maine CDC said the same.

"So far at least, this looks like an H1N1 dominated year, and the years where H1N1 dominates, you can have very large outbreaks, but the number or the proportion of people … who get very sick, is relatively small compared to years that are dominated by H3N2."

Brian Ward, professor emeritus, Research Institute of the McGill University Health Centre, to CBC News

Relatively small is not zero. Week 39 brought 3,334 lab-confirmed flu hospitalizations in the U.S., 1.0 per 100,000 people, and adults 65 and older already have the highest rate, 3.3 per 100,000. Last season, 198 children died of flu in the U.S.

Nor does early guarantee an early peak. In CDC's clinical-lab series, recent seasons peaked between late December and February. The exception, 2022-23, peaked in the week ending Dec. 10, 2022. A precedent, not a forecast, but the scenario worth planning around.

The case against waiting

The case for holding off is real: protection fades. A CDC-led study of waning protection found "an absolute decline in VE of about 8%–9% per month postvaccination." Against H1N1 it was 8.0% per 30 days; among adults 65 and older, about 10% to 11%, lead author Jill M. Ferdinands told Healio.

Run the arithmetic as an illustration, not a prediction. From today, Wednesday, Oct. 14, a peak like 2022's Dec. 10 is 57 days away; at that pace, a shot today would lose roughly 15 to 17 points of effectiveness by then. A peak like Feb. 1, 2025, is 110 days off, roughly 29 to 32 points.

Waiting claws some of that back for February. The cost is protection now, with H1N1 already circulating and immunity taking about two weeks to build: a shot today covers you from about Oct. 28. The study has limits: it measured hospitalized U.S. adults in the 2015-16 to 2018-19 seasons, and its H1N1 estimate carried a wide confidence interval (1.4% to 21.9%).

That waning is why CDC discourages July and August shots, not October ones. Its guidance says vaccination "should ideally be offered during September or October," and for most people "before the end of October." Dr. William Schaffner, a professor of preventive medicine and infectious diseases at Vanderbilt Health, put it more bluntly to Live Science: October is the ideal time to get vaccinated, and even more so right now.

Some people have more reason to move first. CDC says adults 65 and older "should preferentially receive" a high-dose, recombinant or adjuvanted vaccine. Children 6 months to 8 years who haven't had two or more doses before July 1 need two doses at least four weeks apart, so a first dose today means a second no earlier than Nov. 11. Pregnant people can be vaccinated in any trimester; people with asthma, diabetes or heart disease are also at higher risk. Anyone at higher risk who gets sick should call early: antivirals such as Tamiflu work best started within two days of symptoms, and CDC recommends early treatment for them.

In Canada, Quebec opened bookings on Oct. 13, for a campaign starting Oct. 20, free only for high-risk groups; others pay. In the U.S., a January 2026 HHS memo dropped universal childhood flu recommendations, while the American Academy of Pediatrics and other groups issued their own guidance that everyone 6 months and older get a yearly shot, USA Today reported; CDC says its July 2025 flu recommendations remain in effect.

The trade is lopsided: a few extra points of protection in a February that may not be the peak, against none at all while H1N1 is already spreading.