key: cord-0983228-44po8vi1
authors: Sarjomaa, M.; Diep, L. M.; Zhang, C.; Tveten, Y.; Reiso, H.; Thilesen, C.; Nordbo, S. A.; Berg, K. K.; Aaberge, I.; Pearce, N.; Kersten, H.; Vandenbroucke, J. P.; Eikeland, R.; Fell, A. K. M.
title: SARS-CoV-2 antibody persistence after five and twelve months: A cohort study from South-Eastern Norway
date: 2022-02-21
journal: nan
DOI: 10.1101/2022.02.16.22271075
sha: bba03e3058987716ade249cf08136e3dd8dfff83
doc_id: 983228
cord_uid: 44po8vi1

Objectives: To assess total antibody levels against Severe Acute Respiratory Syndrome CoronaVirus 2 (SARS CoV-2) spike protein up to 12 months after Coronavirus Disease (COVID-19) infection in non-vaccinated individuals and the possible predictors of antibody persistence. Methods: This is a prospective multi-centre longitudinal cohort study. Participants The study included SARS-CoV-2 real-time polymerase chain reaction (RT-PCR) positive and negative participants in South-Eastern Norway from February to December 2020. Possible predictors of SARS-CoV-2 total antibody persistence was assessed. The SARS-CoV-2 total antibody levels against spike protein were measured three to five months after PCR in 391 PCR-positive and 703 PCR-negative participants; 212 PCR-positive participants were included in follow-up measurements at 10 to 12 months. The participants completed a questionnaire including information about symptoms, comorbidities, allergies, body mass index (BMI), and hospitalisation. Primary outcome The SARS-CoV-2 total antibody levels against spike protein three to five and 10 to 12 months after PCR positive tests. Results: SARS-CoV-2 total antibodies against spike protein were present in 366 (94%) non-vaccinated PCR-positive participants after three to five months, compared with nine (1%) PCR-negative participants. After 10 to 12 months, antibodies were present in 204 (96%) non-vaccinated PCR-positive participants. Of the PCR-positive participants, 369 (94%) were not hospitalised. The mean age of the PCR-positive participants was 48 years (SD 15, range 20-85) and 50% of them were male. BMI [≥] 25 kg/m 2 was positively associated with decreased antibody levels (OR 2.34, 95% CI 1.06 to 5.42). Participants with higher age and self-reported initial fever with chills or sweating were less likely to have decreased antibody levels (age: OR 0.97, 95% CI 0.94 to 0.99; fever: OR 0.33, 95% CI 0.13 to 0.75). Conclusion Our results indicate that the level of SARS-CoV-2 total antibodies against spike protein persists for the vast majority of non-vaccinated PCR-positive persons at least 10 to 12 months after mild COVID-19.

Since the initial outbreak of COVID-19 was reported in Wuhan in December 2019, over 304 million 73 people have been infected worldwide, with over 5.4 million deaths reported by the World Health 74

Organization (WHO) as of 11th January 2022 [1] . The first wave of the pandemic in Norway peaked in 75

March 2020. The second wave started in the autumn of 2020 and the third wave in February 2021. 76

The estimated seroprevalence of SARS CoV-2 antibodies in Norway was 0.6% in the late summer of 77 2020 and increased to 3.2% in January 2021, after the second wave [2] .

The impact of SARS-CoV-2 on human health in individuals and populations depends on multiple 79 factors such as the level of healthcare, diagnostics, therapeutics, social distancing measures such as 80 lockdowns, face masks, working from home, and the availability and coverage of vaccines to control 81 the disease. Understanding the cellular and humoral immunity to COVID-19 is necessary to assess the 82 future course of the pandemic. There is still insufficient data regarding the long-term persistence of 83 antibodies and the level of protective immunity, especially in patients who underwent mild infections 84 and those who were not hospitalised [3] [4] [5] [6] [7] [8] [9] . Hence, there is a need for studies on antibody kinetics to 85 improve our understanding of humoral immunity following COVID-19 infections.

The gold standard for antibody test assays has not been determined, and numerous immunoassays 87 have been developed [10] . IgG antibodies against SARS-CoV-2 are highly sensitive markers 7-14 days 88

after symptom onset [8, 11] . Antibodies are detected in 90% of individuals after two weeks and are 89

highly correlated with neutralising antibodies [8, 12] . High sensitivity and specificity are important for 90 all serological assays, and the specificity of an antibody test might be an issue when the infection 91 prevalence is low [13] . Some smaller studies early in the pandemic showed that antibodies declined 92 within a few months after infection [14] [15] [16] . Antibody levels against the nucleocapsid protein have 93 been shown to decline more rapidly than antibodies against the spike protein [12] . Furthermore, the 94 kinetics and protective immunity between anti-nucleocapsid and anti-spike antibodies may differ 95 [17] . Data from large, systematic, and quantitative follow-up studies of antibodies for longer than six 96

to eight months are limited [18, 19] .

The study of antibody persistence is still important, despite vaccination, as antibody longevity during 98 an ongoing pandemic is of scientific interest, as well as being particularly relevant for providing 99 correct and informed vaccination strategies to those who have had an infection. However, there is 100 still a lack of knowledge regarding the longevity of protective immunity.

Our aim was to assess the SARS-CoV-2 total antibody levels against spike protein up to 12 months 102 after COVID-19 infection in non-vaccinated individuals, and identify possible predictors for antibody 103 persistence [20] . is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted February 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 113

Participants who did not consent or were unable to answer a questionnaire (Norwegian language) 114 were excluded. COVID-19 vaccination was introduced in January 2021, and post-vaccine samples 115

were excluded in this study.

The official Norwegian testing criteria for SARS-CoV-2 changed over time but were the same for the 117 matched PCR+ and PCR-participants. In the first wave of the pandemic, PCR testing was restricted to 118 patients with symptoms. In the second wave, PCR testing was also applied to close contacts and 119 asymptomatic persons during outbreaks. All participants were invited to the sampling of antibodies 120

and were asked to fill in the self-reporting questionnaire simultaneously, three to five months ( representative was involved in the process of piloting the questionnaire. Questions are shown in S1 129 Table. 130

The questionnaire consisted of questions about demographic data, income, education, smoking 131 habits, hospitalisation, and comorbidities such as asthma and chronic obstructive pulmonary disease 132 (COPD), other lung diseases, cancer, heart disease, diabetes, hypertension, musculoskeletal disease, 133 any other disease and pollen allergy. Questions about symptoms included the presence or absence of 134 cough, runny nose, stuffy nose, sore throat,dyspnoe, headache, fever with chills or sweating, 135 abdominal pain, nausea, diarrhoea, impaired sense of smell and taste, myalgia, and dizziness. 136

Questions regarding fatigue and reinfection were also included.

Self-reported vaccination data were available and checked against data from the National 138

Immunisation Register (SYSVAK). Additional demographic data (age, sex, and time and place for 139 SARS-CoV-2 PCR) were also registered.

Venous blood samples were obtained at recruitment, three to five months after the PCR test (T1), 142 and again after 10-12 months (T2). All serum samples were prepared from whole blood following 143 centrifugation for 10 min at a minimum of 1800 g at room temperature and stored at -80 0 C until 144 further analysis. Total immunoglobulin levels were analysed at Telemark Hospital using the Siemens 145

Advia Centaur XP SARS-CoV-2 Total assay for the qualitative detection of total antibodies (IgM and 146

IgG) in human serum. On a large panel of blood samples, the Siemens assay achieved a sensitivity 147

and specificity of at least 98% [10]. The assay is a fully automated 1-step antigen sandwich 148 immunoassay using acridinium ester chemiluminescent technology and recombinant SARS-CoV-2 S 1 149 receptor binding domain (RBD) antigen. The total antibody ranged from 0 to 9.99, and ≥ 10 was the 150 upper limit of the assay. The threshold for reactivity was ≥ 1.0 Index. To make it possible to interpret 151 changes in antibody levels, antibody levels were categorised into four categories: negative (0-0.79), 152

low ( is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Statistical analysis 163 Mean, standard deviation (SD), medians, interquartile ranges (IQR), and minimum and maximum 164 values were reported for continuous data as appropriate. Categorical data were reported as 165 frequencies and percentages.

The difference between the observed antibody values at T1 and T2 were calculated. The changes that 167 were greater than the maximum level of 10 Index could not be quantified due to the limitations of 168 the assay. Hence, for further analysis, the participants were categorised into four antibody groups: 169 decreased, unchanged, increased, and maximum level ≥ 10. An increase or decrease of at least 20% 170

in antibody values from T1 to T2 was defined as a significant change; otherwise, the participant was 171 assigned to the unchanged antibody group. Chi-squared and Fisher's exact tests were used to 172 examine differences in proportions between the decreased antibody group and the other groups 173

(unchanged, increased, and maximum antibody level ≥ 10) for binary variables. For continuous and 174

ordinal variables, the Kruskal-Wallis test and the One-Way Anova was used to study the difference. 175

Total symptom and comorbidity scores were calculated by adding the number of symptoms or 176 comorbidities for each participant, and used instead of single variables in the multivariate analysis.

Logistic regression models were used to study the possible predictors for changes in antibody levels. 178

Univariate analysis was performed for each predictor, adjusting for age and sex. Then, a multivariate 179 analysis including all predictors was performed, and the standard error of the effect estimate was 180 compared to that of the corresponding univariate analysis for each predictor to check for collinearity. 181

Odds ratio (OR) was reported with 95% confidence intervals (CI) for association between decreased 182 antibody level and predictors. 

The demographics of the study population by baseline are presented in Table 1 . is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 198 However, PCR+ participants reported more symptoms such as dyspnoea, headache, fever with chills 199 or sweating, abdominal pain, nausea or diarrhoea, impaired sense of taste and smell, myalgia, and 200 dizziness compared to the PCR-participants. In contrast, PCR-participants reported more nasal 201 symptoms (runny/stuffy nose), sore throat, and pain upon swallowing compared to the PCR+ 202 participants. Coughing was similar in both groups. Self-reported moderate or serious fatigue was 203 more frequent among PCR+ than PCR-participants. Only 22 (6%) of the PCR+ participants were 204 hospitalised due to COVID-19. There were no self-reported reinfections during the study period.

The duration and changes in total antibody levels against the SARS-CoV-2 spike protein are shown in 206 Table 2 .

207 208 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint (53%) had values above the upper limit of quantification of the total antibody at both measurements. 219

The time between PCR and antibody measurements is shown in S2 Table. 220

The distribution of variables in PCR+ participants with decreased, unchanged, increased and antibody 221 levels ≥ 10 at time points T1 and T2 are shown in Table 3 . is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Univariate and multivariate regression analyses were performed to assess the possible predictors for 232 antibody levels after 10-12 months, and to compare the group with decreased antibody levels with 233 the other groups (increased, unchanged, and antibody ≥ 10), as shown in Table 4 . is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Among the PCR negative participants, nine had total antibodies ≥ 0.8 at T1, and after initial 249 symptoms, they did not report reinfection between the time of the PCR test and the antibody 250 measurement. The most likely explanation is that they had COVID-19 and a false-negative PCR test, a 251 test taken incorrectly or a low virus level at the time of the test.

We did not detect any self-reported reinfections during the study period. A recent large prospective 253 multicentre cohort study from all the regions in England showed an 84% lower risk of reinfection 254 after seven months of follow-up post SARS-CoV-2 infection among healthcare workers (HCWs) with 255 positive antibodies tested with a range of assays [25] . The correlation between antibody levels and 256

protective immunity remains unclear.

The results for antibody persistence after 10 to 12 months in our study were in accordance with 258 those of several studies with a shorter follow-up time [9, 19, 21] . A seroprevalence study from 259

Iceland reported that pan-IgG antibodies did not decline within four months after the diagnosis of 260 SARS-CoV-2 infection in ( is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted February 21, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 

BMI ≥ 25 kg/m 2 was strongly associated with decreased antibody level in our study. This was not in 297 agreement with a recent Norwegian study that showed the severity of initial illness, older age, and 298

higher BMI were independently associated with increased antibody levels two months after infection 299 [34] . A large prospective cohort study from the United States showed homogenous immune activity 300 across BMI categories [35] . A large retrospective study from Israel showed that the peak level of 301 neutralising antibodies was associated with obesity, with the highest level in patients who were 302 severely obese [36] . The study did not include any follow-up antibodies. A study from Turkey showed 303 no effect of BMI on antibody titres, but the sample size was small and the follow-up period was only 304 60 days [37] . Nevertheless, obesity was a risk factor for COVID-19 infection severity in many studies 305 as shown in a meta-analysis of 42 studies [38] . It is possible that the diversity of the results of the 306 association between BMI and antibody persistence after COVID-19 infection is dependent on disease 307 severity.

There was no association between gender, income, education, and antibody persistence in our study. 309

Socioeconomic differences in Norway are relatively small, and most of the included participants were 310 from the middle class and above, as reflected in the high levels of education and income among the 311 participants.

Our study has several strengths and limitations. A strength of our study is the relatively large and 313

unselected study cohort recruited during the first and second pandemic waves in Norway. The time 314

for the antibody test was related to the PCR test, and not the onset of symptoms. Another strength is 315 the longitudinal design with a relatively long follow-up time. Validated serology assays were used, 316 and the participants answered the questionnaire on the same day as the serum sampling. 317 318

A limitation of our study is that changes in antibody levels above the assay's upper limit value 10 319

Index could not be determined, but it is probably in the group with low antibody levels that 320 significant changes will appear. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Eikeland Myrnes and Siri Cathrine Rølland for their essential assistance for data collection and 356

analysis. The authors would like to express their gratitude to all the participants in this study and to 357 the patient representatives. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted February 21, 2022. ; https://doi.org/10.1101/2022.02.16.22271075 doi: medRxiv preprint

Weekly epidemiological Update and Weekly Operational Update. 361 WHO

Antibody response 367 to SARS-CoV-2 infection in humans: A systematic review

Recovery 370 scenario and immunity in COVID-19 disease: A new strategy to predict the potential of reinfection

How can we interpret SARS-CoV-2 antibody test results? Pathog Dis

Antibody 375 response After SARS-CoV-2 infection and implications for immunity: A rapid living review

Dynamics of 378 neutralizing antibody titers in the months after SARS-CoV-2 infection

Persistence and decay of 381 human antibody responses to the receptor binding domain of SARS-CoV-2 spike protein in COVID-19 382 patients

National SARS-CoV-2 Serology Assay Evaluation Group. Performance characteristics of five 387 immunoassays for SARS-CoV-2: a head-to-head benchmark comparison

Antibody responses to SARS-CoV-2 390 in patients with COVID-19

The duration, 392 dynamics and determinants of SARS-CoV-2 antibody responses in individual healthcare workers

Seroepidemiologic 395 study designs for determining SARS-COV-2 transmission and immunity

Orthogonal 398 SARS-CoV-2 serological assays enable surveillance of low-prevalence communities and reveal durable 399 humoral immunity

Clinical and immunological assessment 401 of asymptomatic SARS-CoV-2 infections

Rapid 404 decay of anti-SARS-CoV-2 antibodies in persons with mild Covid-19

407 Longitudinal follow-up of IgG anti-nucleocapsid antibodies in SARS-CoV-2 infected patients up to 408 eight months after infection

Antibody 410 persistence in the first 6 months following SARS-CoV-2 infection among hospital workers: a 411 prospective longitudinal study

CoV-2 assessed for up to 8 months after infection

Robust neutralizing 416 antibodies to SARS-CoV-2 infection persist for months

Antibody levels in 419 a cohort of pregnant women after the 2009 influenza A(H1N1) pandemic: waning and association 420 with self-reported severity and duration of illness. Influ Other Respir Viruses

Association of respiratory symptoms and asthma with occupational exposures: findings from a 424 population-based cross-sectional survey in Telemark

Excess risk and clusters of symptoms after COVID-19 in a 427 large Norwegian cohort

Association between ChAdOx1 nCoV-19 429 vaccination and bleeding episodes: large population-based cohort study

SARS-CoV-2 infection rates 432 of antibody-positive compared with antibody-negative health-care workers in England: a large, 433 multicentre, prospective cohort study (SIREN)

Humoral immune response to SARS-CoV-2 in Iceland

Prevalence and 439 Durability of SARS-CoV-2 Antibodies Among Unvaccinated US Adults by History of COVID-19

Antibody 442 response to SARS-CoV-2 in infected patients with different clinical outcome

SARS-CoV-2 immune responses for six months based on the clinical severity of COVID-19

-2 induces a durable and antigen specific humoral immunity after asymptomatic to mild 449 COVID-19 infection

Time series analysis and 452 mechanistic modelling of heterogeneity and sero-reversion in antibody responses to mild SARS-CoV-453 2 infection

455 Longitudinal follow-up of IgG anti-nucleocapsid antibodies in SARS-CoV-2 infected patients up to 456 eight months after infection

Dynamics of SARS-CoV-2 458 neutralising antibody responses and duration of immunity: a longitudinal study

Long COVID in 461 a prospective cohort of home-isolated patients

Epidemiological 464 and Immunological Features of Obesity and SARS-CoV-2

The association 466 between obesity and peak antibody titer response in COVID-19 infection. Obesity (Silver Spring)

Analysis of long-term antibody 469 response in COVID-19 patients by symptoms grade, gender, age, BMI, and medication

Mortality-related risk factors of COVID-19: a systematic review and 472 meta-analysis of 42 studies and 423,117 patients

Asthma and asthma-like symptoms in adults assessed by 475 questionnaires. A literature review

How to 477 interpret and use COVID-19 serology and immunology tests

The potential clinical 480 utility of measuring severe acute respiratory syndrome coronavirus 2-specific T-cell responses

Robust T Cell 483 Immunity in Convalescent Individuals with Asymptomatic or Mild COVID-19

Questions from the questionnaire concerning demographic data, hospitalisation, 489 comorbidities, symptoms and fatigue

The median time between positive PCR and antibody measurements. The median time 492 between positive PCR and T1 antibody measurement was 127 days (91-153) and between positive 493 PCR and T2 antibody 310 days