Sustained seropositivity up to 20.5 months after COVID-19

Abstract

This study evaluated the persistence of IgM, IgA, and IgG to SARS-CoV-2 spike and nucleocapsid antigens up to 616 days since the onset of symptoms in a longitudinal cohort of 247 primary health care workers from Barcelona, Spain, followed up since the start of the pandemic. The study also assesses factors affecting antibody levels, including comorbidities and the responses to variants of concern as well as the frequency of reinfections. Despite a gradual and significant decline in antibody levels with time, seropositivity to five SARS-CoV-2 antigens combined was always higher than 90% over the whole study period. In a subset of 23 participants who had not yet been vaccinated by November 2021, seropositivity remained at 95.65% (47.83% IgM, 95.65% IgA, 95.65% IgG). IgG seropositivity against Alpha and Delta predominant variants was comparable to that against the Wuhan variant, while it was lower for Gamma and Beta (minority) variants and for IgA and IgM. Antibody levels at the time point closest to infection were associated with age, smoking, obesity, hospitalization, fever, anosmia/hypogeusia, chest pain, and hypertension in multivariable regression models. Up to 1 year later, just before the massive roll out of vaccination, antibody levels were associated with age, occupation, hospitalization, duration of symptoms, anosmia/hypogeusia, fever, and headache. In addition, tachycardia and cutaneous symptoms associated with slower antibody decay, and oxygen supply with faster antibody decay. Eight reinfections (3.23%) were detected in low responders, which is consistent with a sustained protective role for anti-spike naturally acquired antibodies. Stable persistence of IgG and IgA responses and cross-recognition of the predominant variants circulating in the 2020-2021 period indicate long-lasting and largely variant-transcending humoral immunity in the initial 20.5 months of the pandemic, in the absence of vaccination.

Keywords: Antibody; COVID-19; Duration; Health care workers; IgA; IgG; IgM; Kinetics; SARS-CoV-2; Seroprevalence.

Fig. 1

SARS-CoV-2 seropositivity in a cohort of pre-exposed non-vaccinated health care workers over 2020 and 2021. SARS-CoV-2 IgA, IgG, and IgM antibody (Ab) levels (log₁₀ median fluorescence intensity, MFI) by days since COVID-19 symptoms onset. Black dots represent seropositive and gray ones represent seronegative responses. Samples from the same participant are joined by gray lines. Highlighted in red are samples from individuals after a documented reinfection by RT-qPCR. The blue solid line represents the predicted population average calculated using linear mixed models with linear and quadratic fix effect terms for the dependency on time since symptoms onset. Dashed lines correspond to 95% confidence interval. Predicted antibody level changes relative to levels at the onset of symptoms are reported in the table below at 300 and 600 days after it. Reported marginal R² gives a measure of the goodness of fit and corresponds to the ratio of variance explained by time since infection over the total variance of the outcome, including the modeled random intercept. Significance of fits departing from that of lack of antibody change (null hypothesis) were assessed using a log-likelihood ratio test comparing a full model containing a linear and quadratic term for time since infection and a reduced model containing none of them. Ab, antibody