A child, a teenager and an adult enter a household… No, this is not the beginning of a joke. This is the beginning of a question for you. So, a child, a teenager and an adult enter a household, who do you think would be most likely to infect others in the household? If I were to venture a guess, I would say the child would be most likely to infect others in that household. The child has the most underdeveloped immune system among the three after all, and I have frequently heard the ‘must have been because of the daycare’ rhetoric from young parents who fall sick during the flu season. However, I would be wrong, at least if it’s the SARS-CoV-2 virus.
In their recent article, Christiaan H. van Dorp, a postdoctoral researcher at Columbia University Medical Center, and his colleagues find that adults are most likely to introduce an infection into a household, closely followed by teenagers, whereas children were only ~60% as likely as adults to introduce an infection into a household. They estimated that the rate at which an adult introduced an infection into a household was 0.0008 per day in mid-October 2020, which means that a group of 1200 adults infect one uninfected person every day.
The authors of this article used a prospective household study rather than a reactive household study. A reactive household study is not suitable to estimate how often an infection is introduced into a household as only households that are already infected are included in the study and the time of the infection is unknown. A prospective household study, in contrast, follows all participating households semi-passively during the at-risk period. In the Dutch study that Christiaan and team used for their analysis, more intensive follow-ups were performed upon notification of acute respiratory symptoms in the household.
The participating households were tracked for a maximum of 161 days. The team found that SARS-CoV-2 was introduced and established in 59 out of 307 households (~19%) between August 2020 and March 2021. The surge in hospitalization of hundreds of people in the same period closely followed the number of household infections (Figure 1). The peaks in hospital admissions (see the yellow dots) were always 1-2 weeks after the peak in the household infection rate.
Figure 1: The number of infections in the households is closely followed by the number of hospitalizations reported. The blue line is a measure of the infection rate of a household and the gray region around it is the variation in that rate.
The Netherlands was under lockdown during the period of the prospective household study. Christiaan and his colleagues’ analysis did not test whether children were less likely to introduce the infection into a household because of their limited interaction with the outside world during the lockdown. However, they tested whether the age of a person was a factor in how fast the virus was transmitted within a household upon its introduction into the household. They categorized the population into three different age groups: children (0-12 years old), adolescents (12-18 years old) and adults (above 18 years old). The team formulated a ‘full within household model’ where they tested all possible interactions between the three different age groups (see Figure 2). They found that once an infection had been introduced into a household, children were the primary age group that transmitted the infection within the household. The transmission rates due to adolescents and adults within a household were at most half of that of children, the rates for adolescents being marginally higher than that for adults (Figure 2).
Figure 2: All possibilities of transmission among the three age groups were tested in the full within household transmission model.
Using the full within household model again, Michiel van Boven, a researcher at University Medical Center Utrecht in the Netherlands, tested the effectiveness of a vaccine in controlling the spread of infection within a household. He assumed that a vaccine is ‘leaky’, which means that a vaccinated individual who contracts the virus would not be infected, however, can still spread the virus to others in the household. The simulations showed that vaccinating just the adults was effective in controlling the infection in the household. Vaccinating the adolescents, on top of the vaccination of adults, did not improve the control of infection. It is surprising that although adolescents are likely to introduce the infection into a household and also transmit it within the household, their vaccination does not improve the control of infection.
Christiaan and the team’s results from the prospective household study, therefore, shows that an adult brings the SARS-CoV-2 virus home which is then primarily transmitted within the household by children. Perhaps I was partially correct. A child would be most likely to infect the household, if it is already infected, otherwise it’s the adult who is most likely to infect the household.
Reviewed by: Christiaan van Dorp, Erin Cullen, Trang Nguyen, and Giulia Mezzadri