Abstract
Background
Longitudinal serological surveillance is critical to understand dynamics of SARS-CoV-2 infections in children, a substantial portion of whom are asymptomatic. We describe trends in seroprevalence using at-home testing and evaluate demographic and clinical characteristics associated with seroconversion.
Methods
Children 2–17 years old enrolled at 3 North Carolina sites (Figure 1) were followed April 2- December 31, 2021. Daily electronic surveys solicited symptoms and vaccination status. Four fingerprick lateral flow immunoassay tests were shipped to participants to be completed monthly; sensitivity and specificity were 84.5% and 99.0%, respectively.
We defined an infection window as 30 days before a positive antibody test (IgG), excluding results after any vaccine. Asymptomatic was defined as absence of “new” symptoms in the window; “new” was defined as not occurring the 14 days preceding the first observation of the symptom in the window. Univariate logistic regression was used to compare participants with and without infection-induced antibodies. For estimated seroprevalence, we used Bayesian inference accounting for sensitivity and specificity, modeling IgG positives from a binomial distribution.
Figure 1.
Participants' Counties of Residence in North and South Carolina
Results
Of 1,501 participants, 9.1% developed infection-induced antibody (Table 1). Blacks were more likely to seroconvert (OR 1.95 [95% CI 1.05–3.45]) as were those in a 5-or-more person household (OR 4.25 [CI 1.47–12.1]). Cumulative seroprevalence of SARS-CoV-2 increased from 12.7% in May to 32.4% in October (Figure 2); 61% of those seropositive were asymptomatic (Table 2). Adolescents had the highest seroconversion rates and were significantly more likely than 2–4-year olds to be asymptomatic.
Table 1.
Demographic and Household Characteristics of Participants with and without Infection-induced Antibodies.
Figure 2.
Cumulative Seroprevalence in Unvaccinated Children and Adolescents, May to October, 2021
Bayesian seroprevalence estimates accounting for assay sensitivity and specificity. Error bars represent 95% credible intervals.
Table 2.
Proportion of Participants with and without Symptoms Prior to Seroconversion
Conclusion
Prior SARS-CoV-2 seroprevalence data in children are limited by cross sectional design and use of convenience samples. With serial testing, we demonstrate rising SARS-CoV-2 seroprevalence, with highest rates of seroconversion in adolescents. Although prior findings suggest that asymptomatic infections occur most frequently in young children, we found a high proportion among adolescents. Our findings underscore the importance of serosurveillance to optimize public health efforts aimed at children and adolescents.
Disclosures
Roberto P. Santos, MD, MSCS, American Board of Pediatrics - Sub-Board Infectious Diseases: Honoraria|Eli Lilly and Company: Grant/Research Support|Eli Lilly and Company: Eli Lilly and Company awarded to the University of Mississippi Medical Center with Roberto P. Santos as the Site PI|Health Resources and Services Administration (HRSA): Grant/Research Support|Health Resources and Services Administration (HRSA): Program Director & Principal Investigator|Janssen Pharmaceutical Companies of Johnson & Johnson (JNJ): Grant/Research Support|Janssen Pharmaceutical Companies of Johnson & Johnson (JNJ): Janssen Pharmaceutical Companies of Johnson & Johnson (JNJ) awarded to the University of Mississippi Medical Center with Roberto P. Santos as the Site|Merck Sharp & Dohme (MSD): Grant/Research Support|Merck Sharp & Dohme (MSD): Merck Sharp & Dohme (MSD) awarded to the University of Mississippi Medical Center with Roberto P. Santos as the Site PI.