Assessment of Clinical and Virological Characteristics of SARS-CoV-2 Infection Among Children Aged 0 to 4 Years and Their Household Members

Key Points Question Do community-acquired SARS-CoV-2 infections differ in adults and children aged 0 to 4 years with respect to incidence, symptoms, and detected viral load? Findings In this cohort study of 690 participants from 175 households in Maryland conducted from November 2020 to October 2021, 54 incident SARS-CoV-2 infections were detected in 8.6% of children aged 0 to 4 years, 11.0% of children aged 5 to 17 years, and 6.3% of adults. Children were more frequently asymptomatic or mildly symptomatic than adults; highest detected viral loads correlated with the number of symptoms in adults but not in young children. Meaning This study’s findings suggest that symptomatic screening for SARS-CoV-2 infection may be insufficient to control outbreaks in settings in which young children congregate.


Introduction
The clinical and epidemiological characteristics of SARS-CoV-2 infection in children have been gradually elucidated during the COVID-19 pandemic. 1 Although multiple studies have estimated incidence rates of pediatric infections and hospitalizations, 2,3 described clinical features, 4 and documented transmission of SARS-CoV-2 by children, [5][6][7] most have examined children younger than 18 years collectively or have disproportionately focused on school-aged children and adolescents. [2][3][4][5][6][7][8] Young children may differ from older children in their clinical presentation and immune response to SARS-CoV-2 and their social interaction patterns, each of which may have implications for disease burden. The burden of pediatric COVID-19 has become increasingly apparent with the Delta and Omicron variants. [9][10][11] Children aged 0 to 4 years are the final age group for which COVID-19 vaccines have been made available in the US. Improved understanding of the epidemiological characteristics of SARS-CoV-2 infections in young children will help facilitate vaccine implementation in this age group.
The SARS-CoV-2 Epidemiology and Response in Children (SEARCH) longitudinal household cohort study was designed to examine SARS-CoV-2 epidemiological features among children aged 0 to 4 years and their household members using intensive routine weekly molecular and serological surveillance in both healthy and symptomatic individuals. The SEARCH study was conducted when SARS-CoV-2 ancestral lineages not considered variants of interest or variants of concern (non-VOI/VOC) were circulating, followed by Alpha and Delta variant lineages. This study assessed the incidence, clinical and virological characteristics, and symptoms of SARS-CoV-2 infections and examined the correlations between SARS-CoV-2 viral load and symptoms by age group and between viral load and SARS-CoV-2 lineage.

Household Identification, Recruitment, and Enrollment
for Disease Control and Prevention approved the study based on the review of the Johns Hopkins University Bloomberg School of Public Health. Adults from eligible households provided written informed consent for themselves and for children younger than 7 years. Children provided written assent (in addition to parental consent) if they were aged 7 to 17 years. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.

Data and Specimen Collection
Study questionnaires were completed electronically by adults and parents of children (with age-appropriate input from children). Enrollment questionnaires included questions about sociodemographic characteristics, social and health history, previous diagnosis of COVID-19 by a health care professional, and in-person attendance at a childcare facility, school, or work. Adults were asked about teleworking and close contact with others (<6 feet) at in-person work. For each household, a single reporter completed a questionnaire about household characteristics.
Custom-designed anterior nasal swab collection kits containing viral transport medium were provided at enrollment; instructions for collection, packaging, and shipment on cold packs were provided by televisit and video. For 8 months, participants completed weekly symptom questionnaires that asked whether they had developed fever, cough, shortness of breath or difficulty breathing, chills, sore throat, diarrhea, muscle aches, change in taste or smell, or other symptoms during the past 7 days. Participants also submitted weekly self-collected nasal swab specimens regardless of illness status; adults collected nasal swab specimens from young children. Participants collected an additional nasal swab if they developed any of the specific symptoms listed previously. Symptomatic participants (or parents of symptomatic children aged <18 years) completed surveys that included a longer list of respiratory and systemic symptoms. 3 Solicited symptoms in participants 2 years and older included fever, feverishness, or chills; muscle or body aches; joint pain; change in taste or smell; headache; abnormal fatigue; nasal congestion or runny nose; sore throat; cough; shortness of breath or difficulty breathing; chest pain; abdominal pain; diarrhea; nausea or vomiting; and eye redness or rash. Solicited symptoms in children younger than 2 years included fever or feverishness, abnormal fatigue, nasal congestion or runny nose, sore throat, cough, shortness of breath or difficulty breathing, diarrhea, vomiting, eye redness or rash, and fussiness or inconsolable crying. Information on receipt of COVID-19 vaccines was solicited monthly. Serum specimens were obtained at enrollment and at approximately 4 months and 8 months after enrollment. A comprehensive summary of data elements and specimens collected is provided in eFigure 2 in Supplement 1. Data were collected and stored using Research Electronic Data Capture (REDCap) tools 12 hosted at Johns Hopkins University.

Antibody Assays
Sera were tested for antibodies against the nucleocapsid and spike protein receptor binding domain of wild-type SARS-CoV-2 using 2 immunoassays (Elecsys-N for the nucleocapsid protein and Elecsys-S for the spike protein; Roche Diagnostics). 13 Both assays were highly concordant in unvaccinated individuals. Further details are provided in eMethods in Supplement 1. Results from the nucleocapsid antibody assay were used to define serologically confirmed infections because vaccination with the COVID-19 vaccines used in the US does not induce antibody to the nucleocapsid protein of SARS-CoV-2. a cycle threshold value lower than 30 were further characterized by quantitative RT-PCR (qRT-PCR) assays at the Wisconsin State Laboratory of Hygiene. 14 The limit of detection by qRT-PCR was 1.0 log 10 copies/mL. For assessment of the correlation between symptoms and viral load, specimens with qualitative RT-PCR cycle threshold values greater than 30 that precluded testing by qRT-PCR were assigned a value of 0.2 log 10 copies/mL.

Analytic Definitions
We defined SARS-CoV-2 infection that occurred during the surveillance period as (1) having 1 or more nasal swab specimen with a positive result for SARS-CoV-2 during the surveillance period or (2) having SARS-CoV-2 nucleocapsid antibody seronegative results at enrollment and nucleocapsid antibody seropositive results at 4 months or 8 months after enrollment. SARS-CoV-2 infections were categorized as serologically confirmed if they met the second definition only. Fully vaccinated individuals had completed a primary COVID-19 vaccine series 2 or more weeks before the indicated time point (eTable 2 in Supplement 1). Study completion was defined as participation with fewer than 4 missed weeks of reporting and nasal swab sample collection. Participant-reported underlying medical conditions were categorized based on whether available data suggested that the condition conferred an increased risk of experiencing severe COVID-19 (eg, asthma, obesity, or overweight) (Table). 1-4.9] log 10 copies/mL), data from both groups were aggregated. A sensitivity analysis of the correlation between highest detected viral load in routine specimens and number of symptoms excluding specimens with qRT-PCR results that were assigned a value of 0.2 log 10 copies/mL was also performed.

Statistical Analysis
Fisher exact, Wilcoxon rank sum, and Spearman rank correlation tests were used as appropriate.
Tests were 2-tailed, and P < .05 was considered statistically significant. Analyses were conducted using R software,        log 10 copies/mL). However, the correlation between median highest detected viral load and number    Age group, y B, Solicited symptoms in participants 2 years of age and older included fever, feverishness, or chills; muscle or body aches; joint pain; change in taste or smell; headache; abnormal fatigue; nasal congestion or runny nose; sore throat; cough; shortness of breath or difficulty breathing; chest pain; abdominal pain; diarrhea; nausea or vomiting; and eye redness or rash. Solicited symptoms for children younger than 2 years included fever or feverishness, abnormal fatigue, nasal congestion or runny nose, sore throat, cough, shortness of breath or difficulty breathing, diarrhea, vomiting, eye redness or rash, and fussiness or inconsolable crying.

SARS-CoV-2 Infections
of illness symptoms differed between adults and children. Among adults (n = 20), the median highest detected viral load was significantly correlated with the number of symptoms when qRT-PCR values and the 0.2 log 10 copies/mL assigned value for RT-PCR-positive unquantified specimens were analyzed (R = 0.69; P < .001) ( Figure 4C)  log 10 copies/mL; P = .73).

Discussion
The SARS-CoV-2 viral load may be correlated with infectiousness. 17 In this study, we found that SARS-CoV-2 viral load and number of symptoms were highly correlated in adults, but there was no correlation between viral load and number of symptoms among children in both age groups. If viral load as measured by qRT-PCR testing is an accurate sign of infectiousness, our findings may have implications for the mitigation of SARS-CoV-2 infection, particularly among children aged 0 to 4  years. Symptom screening alone is unlikely to be sufficient to identify and stop SARS-CoV-2 outbreaks in congregate settings, such as childcare facilities. Instead, testing of asymptomatic contacts of persons with SARS-CoV-2 infection may remain important for identifying asymptomatic individuals who might be just as likely to transmit SARS-CoV-2. Our findings also suggested that viral load in the upper respiratory tract may reflect the extent of systemic viral replication in adults but not children. Some [18][19][20] but not all 21 studies have found lower viral loads in asymptomatic children and adults than in symptomatic individuals. In contrast, we found no difference in viral loads between asymptomatic and symptomatic young children. The differences between our study results and previous findings may be explained by our use of quantitative (qRT-PCR testing) rather than qualitative (cycle threshold value) methods 22 and by our intensive surveillance methods, which may have detected virus earlier in those with asymptomatic infections, 19 when viral load may be higher. 23 Children in both age groups who experienced RT-PCR-confirmed SARS-CoV-2 infection during the study were more frequently asymptomatic than adults. In general, children with symptomatic SARS-CoV-2 infections also experienced fewer symptoms than adults with symptomatic infections, which was consistent with previous observations. 3,18 Although nasal congestion was frequent in all symptomatic participants, 24 children aged 5 to 17 years were less likely than adults to experience fatigue or change in taste or smell (which could not be readily assessed in children aged <2 years).
These findings highlight the challenge of identifying SARS-CoV-2 infections in young children based on symptom screening because children are more likely than adults to be asymptomatic, and SARS-

Strengths and Limitations
This study has several strengths. We focused on infants and preschool children, a relatively understudied population with respect to SARS-CoV-2 infection. Intensive weekly surveillance in both healthy and symptomatic individuals, high adherence, and minimal attrition among cohort participants allowed for sampling throughout infection and complete symptom ascertainment. In addition, viral load was assessed using a validated quantitative PCR assay. The study also captured 3 successive pandemic waves that included circulation of SARS-CoV-2 non-VOI/VOC, Alpha, and Delta variant lineages.
The study also has limitations. As noted in a previous study, 3 individuals who participate in intensive surveillance studies may not represent the general population, and persons of several racial and ethnic backgrounds and persons with low income were underrepresented in the SEARCH cohort; both of these issues may limit the generalizability of some findings. Moreover, the relatively small number of participants with SARS-CoV-2 infection may emphasize or minimize differences between symptomatic and asymptomatic infections. In addition, because the SEARCH study did not include individuals with Omicron variant infection, the applicability of these findings to Omicron and other newly emerging variants of concern will need to be assessed in future studies. The weekly sampling approach was not designed to identify peak viral load titers during SARS-CoV-2 infections. In addition, the quality of nasal swab specimen collection may have varied between self-collected and parent-collected specimens. However, the distribution of sample timing relative to infection onset

Conclusions
In this cohort study, children aged 0 to 4 years with SARS-CoV-2 infection were frequently asymptomatic or mildly symptomatic, and viral loads in their nasal swab specimens did not correlate with illness severity. Although the implications of these findings for household transmission remain to be evaluated, they suggest that SARS-CoV-2 infection may be underrecognized and that symptoms may not reflect infectiousness in young children.