Identifying susceptibility of children and adolescents to the Omicron variant (B.1.1.529)

As the Omicron variant (B.1.1.529) of SARS-CoV-2 drives a new surge in coronavirus disease 2019 (COVID-19) cases globally, increasing proportion of pediatric cases is noteworthy [1]. In the United States (US), where the Omicron variant had been predominant since December 2021, the seroprevalence of infection-induced SARS-CoV-2 antibodies among children aged 0–11 years increased from 44.2 to 75.2% during December 2021 to February 2022, recording the highest increase among all age groups (overall US seroprevalence increased from 33.5 to 57.7%) [2]. In England, pediatric admissions with COVID-19 infections began to rise since December 26, 2021, from an average of 40 admissions per day to 120 per day, a 3-fold rise in 2 weeks [3].

The rise in pediatric cases might be attributed to the elderly-prioritized vaccination strategy against COVID-19 and relatively higher contact rates among school-aged children than adults. As of May 31, 2022, the World Health Organization (WHO) authorized COVID-19 vaccines for individuals aged 18 years and older, and only one vaccine could be used for individuals from 5 years of age, which let children more vulnerable to COVID-19 [4]. Otherwise, children might genuinely be more susceptible to contracting the Omicron infections than adults. Identifying the age-specific susceptibility to SARS-CoV-2 is of much interest for effective public health strategies and vaccination policy. However, it is often not easy to clarify the age-specific susceptibility to an infection due to the lack of sufficient data.

In South Korea, the biggest 5th wave has been driven by the Omicron variant since January 2022, following the 4th wave by the Delta variant and the 3rd wave by the original SARS-CoV-2 virus (Fig. 1A). Here, we attempted to identify the age-specific susceptibility to the Omicron variant compared to the Delta and the pre-Delta strains, using the epidemiologic data of those three waves along with vaccine coverage data in South Korea. The force of infection (λ_i) experienced by age group i was used to estimate the age-specific susceptibility in this study [5]. Given that South Korea has the National Infectious Disease Surveillance System (NIDSS) which mandates the web-based reporting of every COVID-19 case and the National Immunization Registry, it could be a suitable country to elucidate this topic [6, 7]. Considering the concordance between the laboratory-confirmed COVID-19 incidence and the national seroprevalence of SARS-CoV-2-specific anti-nucleocapsid (anti-N) antibodies (Fig. 1B), the ascertainment ratio in South Korea may not be far from reality [8].

We adapted a previously described model which estimated the age-varying susceptibility to the Delta variant [9] and updated the model with recent vaccine coverage data and waning of vaccine effectiveness against the Omicron infection.

The age distribution of COVID-19 cases during the 3rd (pre-Delta), 4th (Delta), and 5th (Omicron) waves in South Korea is shown in Fig. 2A, B. The proportions of COVID-19 cases among those aged 19 years or less were 11.02%, 16.72%, and 28.55% during the 3rd, 4th, and 5th waves, respectively. Meanwhile, the proportions of cases among those aged 60 years or more were 28.47%, 10.16%, and 10.04% during the 3rd, 4th, and 5th waves, respectively. Considering the age demographics (age skewed older in South Korea), the normalized proportions of cases by age structure among those aged 19 years or less were 13.28%, 23.43%, and 36.95% during the 3rd, 4th, and 5th waves, respectively.

The estimated age-specific susceptibility to COVID-19 is shown in Fig. 2C, D. Both the 3rd (pre-Delta) and 4th (Delta) waves showed a similar age-dependent increase, whereas the 5th (Omicron) wave showed an inverted bell curve with bimodal peaks. A significant difference between the susceptibility to the Omicron and that to the Delta and pre-Delta variants was found in the younger age group. The rise in susceptibility to the Omicron/pre-Delta variant was highest in the 10–15 years age group (5.28 times [95% CI, 4.94–5.60]), whereas in those aged 50 years or more, the susceptibility to the Omicron/pre-Delta remained stable at approximately twofold, scoring the lowest value of 1.12 (95% CI, 1.09–1.14) among those aged 75 years or more. The rise in susceptibility to the Omicron/Delta variant was highest in the 15–19 years age group (3.21 times [95% CI, 3.12–3.31]) and lowest in those aged 75 years or more (0.93 times [95% CI, 0.89–0.97]).

For validation, we simulated the model forward with the estimated susceptibility parameters during the training period and verified the estimated case numbers could reproduce the observed epidemiological patterns equally well. We also presented the convergence of the Markov chain Monte Carlo (MCMC) algorithm through the trace plots and autocorrelation function (ACF) of the parameters. Varying the contact patterns, proportion of asymptomatic cases, and vaccine effectiveness in sensitivity analyses, the rise in susceptibility to the Omicron/pre-Omicron was not changed, recording the higher values in the 0–19 years age groups than in other age groups (Additional file 1: Fig. S1 to S12).

Even after adjusting for contact pattern, vaccination status, and waning of vaccine effectiveness, the age-specific susceptibility among age group 0–19 years was approximately 5 times higher during the 5th wave (Omicron driven) than that during the 3rd wave (pre-Delta) and 3 times higher than that during the 4th wave (Delta driven). Indeed, children are more susceptible to the Omicron variant compared with the previous strains of SARS-CoV-2. According to the US Centers for Disease Control and Prevention report, the hospitalization rates among individuals aged 12–17 years were 3.5 times as high during the peak week of the Omicron period than during the Delta period [27]. Previously mentioned pediatric admissions with COVID-19 infections in England also showed a 3-fold rise in 2 weeks from December 26, 2021 [3]. These findings are in line with our result, although the increased susceptibility to the infection is not necessarily correlated with the increased hospitalization rates.

Considering that the Omicron variant has shifted tropism to the upper respiratory tract from the lower respiratory tract, children whose upper airway is immature and relatively smaller than those in adults could be much more easily affected [28, 29]. Likewise, increasing cases of croup, an acute laryngotracheobronchitis characterized by barking cough, were noted in South Korea, during the Omicron surge [30]. What is more, the endocytic entry which Omicron prefers over the angiotensin-converting enzyme 2 (ACE2)-dependent pathway could also explain the higher number of pediatric cases since children have a lower number of ACE receptors [31, 32]. Taken together, our finding is in accordance with both epidemiological and biological observations.

There are several limitations in this study. First, the contact matrix in this study was not of our own empirical data, but instead mathematically estimated data [23]. Second, although we attempted to reflect the social distancing policy into our model, the implementation of nonpharmaceutical interventions (NPI), which could change the effective contact rates, was not fully considered in this study. However, the essential measures such as mask mandate, contact tracing, and mandatory quarantine for international arrival have been applied consistently throughout the study period (Additional file 1: Table S1). Moreover, the relative increase in susceptibility within the same age group would not change much, given the NPI use was likely to be consistent among the same age group [33]. Third, the exact proportion of asymptomatic infections by age groups remains unclear. To overcome this hurdle, we adopted the result from a prospective cohort study and further conducted a sensitivity analysis with variable ranges of asymptomatic proportions [14]. In addition, public health authorities in South Korea had conducted a mass screening test for those who had contact with confirmed COVID-19 cases until February 2022. For example, if there was one case in a school, whole classmates (occasionally whole students in the school) and their parents were screened for SARS-CoV-2 with either polymerase chain reaction (PCR) or rapid antigen test. Thus, we think it is unlikely to miss a significant number of cases in South Korea.

In conclusion, large-scale testing, prompt epidemiological survey, and vaccination status records in a national registry in South Korea allowed us to analyze the age-stratified susceptibility to SARS-CoV-2. Generally, the Omicron variant of SARS-CoV-2 was estimated to spread more easily among children than the Delta and pre-Delta strains. At the beginning of the SARS-CoV-2 pandemic, an increased number of cases and a greater risk of severe disease with increasing age were notable. We might now see the course of adaptation of novel SARS-CoV-2 to humans. This age affinity seems to be similar with influenza that we include both the youngest and the elderly as target groups for vaccination. Although it is not yet clear whether children could be key driver groups in SARS-CoV-2 transmission hereafter, additional efforts for vaccinating children might be considered to reduce the pandemic’s impact on the whole community.