Strain-Specific Influenza Antibody Waning Emerged During Low-Circulation Years
Reduced influenza circulation during the COVID-19 pandemic raised a straightforward immunologic concern: if people were not being re-exposed to circulating virus, would antibody protection fade at the same pace across strains and populations? A prospective comparative analysis from Michigan and Hong Kong suggests the answer is no. Influenza antibody waning during this period was selective, heterogeneous, and strongly dependent on both strain and setting.
The study drew on two long-running prospective cohorts, the HIVE study in Michigan and EPI-HK in Hong Kong, and focused on participants who had neither influenza infection nor vaccination during the observation window. Serum samples collected from 2020 through 2023 were tested with hemagglutination inhibition assays and microneutralization assays against vaccine strains. Investigators then estimated two-year fold changes in antibody titers using multivariable power-law models adjusted for age and baseline titer. In practical terms, the analysis asked whether the absence of regular viral exposure translated into a broad contraction of measurable humoral immunity, or whether antibody persistence differed by antigenic target.
The answer split along strain lines.
Michigan Cohort Showed Clear Waning for Some H1N1 Strains
In Michigan, 173 sera from 57 HIVE participants were analyzed. Initial HAI geometric mean titers ranged from 12.3 to 123.4. Among participants with baseline titers above 1:40, the clearest waning appeared in H1N1 strains. Estimated fold changes showed declines for A/Brisbane at 0.70, A/Idaho at 0.40, and A/Michigan at 0.43, with confidence intervals that remained below 1.0 for each estimate. Those numbers indicate real contraction in measurable antibody levels over time, not random drift around a stable baseline.
The pattern was not uniform across influenza subtypes. H3N2 and influenza B showed minimal or no waning in the Michigan cohort. That asymmetry matters biologically because antibody durability is not simply a function of elapsed time. It reflects how often memory B cells are restimulated, how stable the antigenic targets remain, and how much prior exposure history has shaped the immune response. During a period of low community circulation, some strains appear to lose serologic ground faster than others, but the immune system does not behave as though every influenza subtype is stored in the same compartment with the same shelf life.
One sentence captures the core result: waning was present, but it was not universal.
Hong Kong Cohort Showed a Different Pattern Across H1N1 and H3N2
The EPI-HK cohort contributed 259 sera from 60 participants. Baseline HAI geometric mean titers ranged from 6.3 to 40.9, lower overall than in Michigan, yet the direction of change still depended on the strain under study. In Hong Kong, H1N1 responses were heterogeneous rather than uniformly declining. Some strains retained more stability than others, while selected H3N2 strains showed more obvious waning.
Among the clearest examples were A/Cambodia, with a fold change of 0.44, and A/Singapore, with a fold change of 0.48. Both estimates point to substantial decline in antibody titers over the two-year period. In a setting where influenza transmission stayed suppressed, this kind of reduction is biologically plausible: without natural boosting from infection, circulating antibody levels can drift downward, especially when the immune response depends on repeated antigenic stimulation to stay elevated.
Even so, the Hong Kong data did not mirror a simple global loss of immunity. Some H1N1 responses remained comparatively stable, reinforcing the main theme of the study: post-pandemic antibody dynamics are strain-specific rather than monolithic.
Microneutralization Results Matched the Hemagglutination Findings
Microneutralization assays largely tracked the same direction as the HAI results. That concordance strengthens the interpretation, since the two assays capture related but not identical features of the antibody response. HAI reflects the ability of serum antibodies to inhibit hemagglutination, while MN provides a functional readout of neutralizing activity. When both shift in the same direction, the signal is less likely to be an artifact of a single test platform.
This concordance is clinically relevant even though serology does not equal protection in a one-to-one manner. Influenza immunity is layered, and antibody titer is only one component of defense. Still, falling titers can signal a narrowing cushion against infection, especially if a strain re-enters circulation after a prolonged period of suppression. The study shows that the immune system retained measurable memory for many strains, but some targets lost more of that serologic reserve than others.
Population Immunity Persisted, but Not in a Uniform Way
The most important interpretive point is also the simplest: the pandemic did not produce a single, global pattern of influenza antibody collapse. Instead, population immunity remained preserved for many strains while waning emerged for selected viruses and differed across Michigan and Hong Kong. Age adjustment and baseline titer adjustment were built into the modeling, which makes the observed differences harder to dismiss as a function of age structure or starting antibody level alone.
The reason this matters at the bedside and in public health planning is plain. If low circulation during pandemic years altered antibody profiles in a strain-dependent way, then the next influenza season would not arrive with evenly distributed susceptibility across subtypes. Surveillance and vaccine strain selection both depend on understanding which parts of the immune repertoire have drifted downward and which remain comparatively intact. In this study, H1N1 waning stood out in Michigan, while selected H3N2 strains were more vulnerable in Hong Kong.
What the Cohorts Can and Cannot Tell Us
The cohorts were relatively small, and the analysis was restricted to people without influenza infection or vaccination during the observation window. Those features sharpen internal validity but limit how broadly the findings can be extended to the general population, where vaccination histories, infections, and repeated exposures are far more complex. The study also focused on antibody titers rather than clinical infection outcomes, so the data describe immunologic change rather than direct protection against future illness.
Even with those limits, the findings have immediate scientific value. They argue against assuming that low influenza circulation produces a uniform collapse in immunity. Instead, different strains aged differently inside the immune memory of two populations separated by geography but linked by the same pandemic-era exposure gap. That distinction is not academic. It is exactly the kind of information that informs how immunologists, epidemiologists, and vaccine planners think about the rebound of influenza after prolonged suppression.
The final message is not that influenza immunity disappeared. It is that influenza antibody waning occurred selectively, with meaningful differences by strain, assay, and region, and those differences should shape how future circulation is interpreted rather than forcing every subtype into a single narrative of decline.