Multiplex serology captured pre-Omicron SARS-CoV-2 immunity across a household cohort
In the months before Omicron spread widely, a Michigan household cohort offered a rare view of SARS-CoV-2 immunity when infection and vaccination still produced more separable antibody histories. Researchers tested serum from participants in the long-running Household Influenza Vaccine Evaluation cohort with three assay platforms that served different purposes at the bedside and in the laboratory: binding IgG measured by multiplex electrochemiluminescence, functional inhibition measured by ACE2 competition, and neutralization measured by pseudovirus assays. The central question was practical and immunologic at once. Could a high-throughput, low-volume platform capture variant-specific antibody activity well enough to support large-scale serosurveillance and age-stratified immunity estimates?
It could, with caveats tied to the biology of each readout.
Strong agreement across assay platforms, with the best performance from binding IgG
The analysis included 156 serum samples from 117 participants. Most samples, 65.8%, came from adults, and 77.8% were collected from people who had been vaccinated during the study window. Within the multiplex electrochemiluminescence platform, IgG measurements moved closely together across antigens, with correlations ranging from 0.95 to 1.00. ACE2 inhibition assays also tracked well with themselves, with correlations from 0.70 to 0.99. Cross-platform agreement was equally notable: ECL results versus pseudoviral neutralization ranged from 0.81 to 0.83, while ACE2 inhibition versus pseudoviral neutralization reached 0.88 to 0.95.
Those numbers matter because they speak to how different laboratory technologies map onto the same underlying immune signal. Binding assays capture the quantity of antibody recognizing viral antigens, while neutralization and ACE2 competition get closer to function by asking whether those antibodies interfere with viral entry. In a fast-moving respiratory virus with multiple spike mutations, a broad assay panel is more useful than a single measurement because strain-specific immunity can shift under pressure from new variants.
Vaccination status was classified well by ECL, while ACE2 results were more uneven
Receiver operating characteristic analysis showed that the electrochemiluminescence IgG assays discriminated vaccination status with an area under the curve of 0.84 to 0.88. Sensitivity ranged from 0.88 to 0.91, and specificity from 0.87 to 0.88. For a population serosurvey, that is a strong performance profile: the assay identified vaccinated participants with high accuracy while keeping false positives relatively low.
The ACE2 inhibition assays told a more variable story. Their area under the curve ranged from 0.53 to 0.96, with sensitivity spanning 0.17 to 0.93 and specificity from 0.91 to 1.00. Lower performance appeared for some BA.2-related antigens. In clinical terms, this is the kind of variability expected when the assay is interrogating a more delicate functional property against closely related viral antigens. Small antigenic shifts in the receptor-binding domain can alter how strongly serum antibodies block ACE2 binding, even when total IgG remains detectable. A binding signal can persist while functional inhibition weakens.
Age shaped the antibody profile, but younger strata remained imprecise
SARS-CoV-2 IgG concentrations generally increased with age and appeared to plateau in early adulthood. The age trend fit the study’s surveillance structure: older participants had accumulated more opportunities for vaccination or infection by the end of 2021, while younger groups were both smaller and more likely to have limited exposure histories. Estimates in younger age groups were imprecise because sample sizes were small, so the age curve should be read as directional rather than definitive.
Age-linked serology in this period reflects more than biology alone. It also mirrors behavior and infection history, because the household cohort had active surveillance for acute respiratory illness and repeated documentation of vaccination and PCR-confirmed infection. In a pre-Omicron setting, those histories were still relatively cleanly separable. Vaccination had not yet been introduced to children, and infection prevalence remained low enough that hybrid immunity was still uncommon. That created a narrow epidemiologic window in which infection-induced and vaccine-induced responses could be studied with less overlap than later in the pandemic.
Functional and binding assays answered different questions about the same immune response
The study design sharpened the distinction between three laboratory approaches. Multiplex electrochemiluminescence was well suited to broad, high-throughput measurement of antibodies to spike, receptor-binding domain, and nucleocapsid antigens from multiple variants, including original strain, Delta, Beta, and Omicron-lineage targets. ACE2 inhibition added a functional layer by measuring whether antibodies blocked the viral interaction with the host receptor. Pseudovirus neutralization remained the most direct functional comparator, but it is labor-intensive and technically demanding, which limits its practicality in large serosurveys.
For population immunity work, the value of the multiplex platform lies in scale. A household cohort with repeated blood collection, routine symptom surveillance, and linked infection history can generate a more complete immune record than a single cross-sectional sample ever could. Here, that infrastructure let investigators compare assay behavior against known vaccination and infection histories rather than treating antibody values as isolated lab outputs.
One finding carries immediate bedside relevance even though the study itself was epidemiologic rather than therapeutic: immune history leaves a measurable fingerprint, and different assays read that fingerprint with different clarity. Binding IgG performed consistently across variants. ACE2 inhibition was informative but less stable for some antigen targets. Neutralization remained the functional reference point, yet the study showed that a multiplex ECL platform can approximate that signal closely enough to support broader surveillance.
Pre-Omicron serosurveillance still depends on careful interpretation of variant-specific signals
The work was conducted before Omicron became dominant, which makes it particularly useful as a baseline for understanding how earlier vaccine- and infection-induced responses behaved under lower levels of antigenic complexity. The investigators found high correlation across ECL, ACE2, and pseudovirus assays, and the age-specific IgG profile aligned with the cohort’s documented vaccination and infection patterns. Those are useful results for surveillance programs that need to infer immunity from limited serum volumes.
The limitations are built into the design rather than hidden from it. Only 156 serum samples were analyzed, and the youngest age groups were too small for precise estimates. ACE2 assay performance was not uniform across all antigens, especially for some BA.2-related targets, which means functional inhibition cannot be assumed to behave identically across variants. Even so, the study provides a practical laboratory framework for tracking pre-Omicron immunity in household populations and for interpreting how variant-specific antibody measures relate to vaccination and infection history.
Data availability restrictions also matter operationally. Individual-level data remain constrained by privacy protections in an ongoing longitudinal cohort, although de-identified data can be requested and the analysis code is publicly available. For future serosurveys, that combination of reproducible code, repeated sampling, and linked clinical surveillance is the strongest part of the design—and the one most likely to support comparisons as newer variants continue to alter the antibody readout.