Prevalence of antibodies against SARS‐CoV‐2 in the Norwegian population, August 2021

Abstract Background One year into the COVID‐19 pandemic, the cumulative number of confirmed COVID‐19 cases in Norway was still low. In January 2021, when the Norwegian COVID‐19 vaccination campaign started, the national seroprevalence estimate of SARS‐CoV‐2 antibodies was 3.2%. We have conducted a nationwide cross‐sectional study in August 2021 to investigate the overall prevalence of SARS‐CoV‐2 antibodies in Norway after 8 months of COVID‐19 mass vaccination and a third wave of SARS‐CoV‐2 infection. Methods Residual sera were collected from laboratories across Norway in August 2021. In IgG antibodies against the spike protein, the spike receptor binding domain (RBD) and the nucleocapsid protein of SARS‐CoV‐2 were measured by a bead‐based flow cytometric assay. Results In total, 1926 residual sera were collected from individuals aged 0–98 years; 55.1% were from women. The overall national estimated seroprevalence from vaccination and/or infection was 62.6% (credible interval [CrI] 60.1%–65.2%) based on having antibodies against both spike and RBD. Estimated seroprevalence increased with age. Among all samples, 11.7% had antibodies against nucleocapsid. For unvaccinated children <12 years, the seroprevalence estimate due to SARS‐CoV‐2 infection was 12.5% (95% CrI 9.3%–16.1%). Of seropositive samples from the unvaccinated children, 31.9% lacked anti‐nucleocapsid antibodies. Conclusions The high overall SARS‐CoV‐2 seroprevalence estimates are in line with Norwegian registry data. Vaccination, not infection, contributed the most to the high seroprevalence in August 2021. Lack of antibodies against nucleocapsid should not automatically be interpreted as absence of previous infection as this could lead to underestimation of COVID‐19 cases in seroprevalence studies.


| INTRODUCTION
Since the first case of coronavirus disease-19  was reported in Norway in February 2020 and until August 2021, there had been three waves of COVID-19 in Norway. 1 However, due to strict non-pharmaceutic interventions, the spread of SARS-CoV-2 infection in the Norwegian population had been quite limited. 2,3 The first COVID-19 vaccine dose was administered in Norway on December 27, 2020, as part of a national mass vaccination campaign. 4 Vaccination was prioritized to front-line healthcare workers, elderly persons, and persons in risk groups, starting at the highest ages and with persons living in retirement homes. 5  Prior to the start of the vaccine campaign, three nationwide cross-sectional seroprevalence studies were conducted in Norway to monitor the development of the pandemic. 7 In January 2021, at the start of the vaccine campaign, only 3.2% (95% credible interval [CrI] 2.3%-4.2%) of the Norwegian population had antibodies against SARS-CoV-2. 7 We wanted to study the progression of SARS-CoV-2 seroprevalence after several months of the national COVID-19 vaccination campaign and a third wave of SARS-CoV-2 infections. We also wanted to investigate whether we could distinguish between antibodies from vaccination or infection. Accordingly, we here present a nationwide cross-sectional seroprevalence study based on residual sera collected in August 2021, that is, after approximately 8 months of COVID- 19

| Antibody analysis
In IgG antibodies against the full-length SARS-CoV-2 spike protein, the spike receptor binding domain (RBD), and nucleocapsid (N) protein were measured in a multiplex bead-based flow cytometric assay as previously described. [9][10][11] Neutravidin-coupled polymer beads with fluorescent barcodes and biotinylated virus proteins were incubated with serum. Subsequently, the beads were stained with R-phycoerythrin-conjugated anti-human-IgG-Fc and analyzed by flow cytometry. The median fluorescence intensity (MFI) measured for each bead subset were divided by the MFI of beads with no antigen (blank). Cut-off-values were based on results from pre-pandemic sera and samples from confirmed COVID-19 convalescents with various F I G U R E 1 Cumulative incidence of confirmed COVID-19-cases and COVID-19 vaccinations in Norway, January-October 2021, and timing of the residual sera collection. Incidence of cumulative infections (in dark blue) from Beredt C19 and frequency of COVID-19 vaccinations (first dose in green and second dose in blue) from the Norwegian Immunisation Registry (SYSVAK). The collection period for sampling of residual sera (July 19-September 12) in the present study is indicated with vertical lines. The shaded area shows the period when 85% of the samples were collected. disease severities. Seropositivity was defined as having antibodies against both spike and RBD (from vaccination and/or infection). In children aged <12 years (unvaccinated individuals), seroprevalence of infection was estimated based on either having antibodies against RBD and spike, or against RBD and nucleocapsid as in previous studies. 7 A subset of the sera from children <12 years was also analyzed using spike S1-protein SARS-CoV-2 IgG ELISA and SARS-CoV-2-nucleocapsid IgG ELISA (Euroimmun, Lübeck, Germany). Borderline samples were considered positive (ratio optical density sample/ internal calibrator ≥0.8).

| COVID-19 cases and COVID-19 vaccinations
Aggregate data on COVID-19 cases were obtained from the Emergency preparedness register for COVID-19 (Beredt C19), which include data from the Norwegian Surveillance System for Communicable Diseases (MSIS). Aggregate data on COVID-19 vaccinations were obtained from the Norwegian Immunisation Registry (SYSVAK). We used data on confirmed COVID-19 cases and vaccinations from July 15, 2021, that is, approximately 3 weeks prior to the major sampling week. We included vaccinations and infections among the population with a valid national identity number and who were registered as residents in Norway in the National Population Registry.

| Statistical methods
Seroprevalence was estimated for Norway and by age groups, sex, and county of residence following the same method as previously described. 7 A Bayesian method that incorporates an uncertain sensitivity and an uncertain specificity of the antibody assay was used for the estimation. 12 Based on validation data from known positive and negative cases, we estimated a sensitivity of 97% (CrI 95%-99%) and specificity of 99.7% (CrI 99.3%-99.9%) when defining seropositivity as having antibodies against RBD and Spike and a sensitivity of 95% (CrI 94%-98%) and specificity of 99.8% (CrI 99.5%-99.9%) when using antibodies against nucleocapsid and RBD. We corrected the overall seroprevalence by county, age group, and sex using a Bayesian multilevel regression post-stratification (MRP) model. 12 The seroprevalence estimates are presented as a point estimate and a 95% CrI. All Bayesian analyses were performed using Stan 13 with the RStan interface. 14 The fraction of positive samples as a function of age was estimated using a generalized additive model with a binomial linkfunction and smoothing spline function with 20 knots for the age variable using the mgcv R-package. 15

| Ethics
The residual sera were irreversibly anonymized at the laboratory of origin. Only aggregated data on confirmed COVID-19 cases and COVID-19 vaccinations were obtained from Beredt C19 and SYSVAK, respectively. The study was approved by the Regional Committee for Medical and Health Research Ethics in Southeastern Norway (reference number 157792). The residual sera were collected from individuals aged 0-98 years. The median age was 29 years, and 55.1% of the sera were from females. Of the sera 385 (20%) were collected from individuals T A B L E 1 Estimated overall seroprevalence and corresponding percentages of COVID-19 vaccinations and confirmed COVID- 19 Figure 2). However, for the age group 18-44 years, the estimated seroprevalence was higher than the percentage of vaccinated individuals, and the estimate was lower than the percentage of vaccinated individuals in the age group ≥65 years (Table 1). A rapid increase in seropositive study samples occurs around the age of 18 years, where the percentage of individuals vaccinated with the first dose also increases rapidly ( Figure 2).

| RESULTS
There were some small differences in the overall seroprevalence estimates from vaccination and/or infection between the 11 Norwegian counties (range 46%-63%) ( Table 2). Children and teenagers were mostly unvaccinated at the time of sampling. Therefore, we also estimated seroprevalence for the different counties based on samples from individuals aged ≥18 years of age ( We observed differences in seroprevalence within the population, mostly according to age, but also according to geographical location.
There was no difference in seroprevalence between males and females.
The overall estimated seroprevalence was highest among the oldest age groups, which was expected, as vaccination was prioritized primarily based on age and given the overall low infection prevalence for all age groups in Norway. 4

| Distinguishing infection from vaccination
We report an estimated prevalence of SARS-CoV-2 antibodies (combination of spike and RBD) of 12.5% (95% CrI 9.3%-16.1%) in unvaccinated children under 12 years. 18 This indicates an actual number of infections that was almost seven times higher than the number of confirmed COVID-19 cases in this age group. However, the seroprevalence estimate due to SARS-CoV-2 infection in the unvaccinated children <12 years may not be directly extrapolated to older age groups because there may be differences in precautions to avoid getting infected and in symptoms and test activity. Moreover, vaccination of older age groups may have offered protection against infection, particularly in the first months after being vaccinated. 5,19 Antibodies against the nucleocapsid protein of SARS-CoV-2 have Some studies have found that children develop lower levels of antibodies against nucleocapsid than adults after SARS-CoV-2 infection. 29 lower rates of seroconversion against nucleocapsid have also been reported in COVID-19 vaccinated individuals with breakthrough infections than after infection of unvaccinated individuals. 21,23,36 Perhaps an earlier control of SARS-CoV-2 by the immune system in children or vaccinated individuals may result in a weaker antibody response against nucleocapsid. 21,[36][37][38][39][40] The antibody assay can also influence detection of anti-N antibodies. 33 Here, lack of nucleocapsid antibodies in some of the seropositive samples from children were confirmed by a second method.
Based on our results and reports by others, we hypothesize that in some individuals, no or low levels of antibodies are induced against the nucleocapsid after SARS-CoV-2 infection. In combination with the documented faster waning of nucleocapsid antibodies, particularly in younger age groups, 23,26,29 this implies that seroprevalence estimates of SARS-CoV-2 infection based on having anti-nucleocapsid antibodies may underestimate the cumulative prevalence of SARS-CoV-2 infection. This may be the case specifically in children, but potentially also in highly vaccinated populations. Therefore, seroprevalence estimates of SARS-CoV-2 infection based on measurements of antibodies against nucleocapsid should be interpreted with caution. This has also been voiced as a concern by others. 21,23,27,33 Nevertheless, if these limitations are taken into consideration, measurements of antinucleocapsid antibodies may contribute valuable information as a supplement to other measurements, as we report here.
The residual sera were collected anonymously. Consequently, it is are collected from a population with potentially higher morbidity than the general population, as well as potential difference in healthseeking behavior. 41,42 However, as an identical sampling strategy has been used several times during the pandemic, 7 the estimates of these studies are comparable over time.
To conclude, we estimate a high overall seroprevalence of SARS-