Speaker
Description
Age-structured heterogeneity in contact behavior, susceptibility, and severity makes epidemic dynamics difficult to interpret from aggregate measures alone. The effective reproduction number $R_e(t)$ collapses this heterogeneity into a single quantity, concealing which subgroups drive transmission and how interventions reshape that balance over time.
We extend elasticity analysis, a method rooted in ecology and population dynamics, to $R_e(t)$. Our framework decomposes $R_e(t)$ into time-varying contributions from each age group, tracking how these shift as epidemic conditions and interventions evolve.
We apply this to age-structured SARS-CoV-2 transmission models fitted to epidemiological data via Bayesian inference, evaluating model outputs across multiple epidemic periods and counterfactual school-based scenarios. In Portugal, adults dominated transmission throughout 2020, yet counterfactual simulations reveal that children and adolescents would have played a decisive role had schools remained open. In the Netherlands, adult dominance was confined to the first lockdown, giving way to adolescents in late 2020 and children in 2021. Across counterfactuals, adolescents consistently exerted stronger early influence than children.
These findings underscore that the dynamical role of any subgroup is not intrinsic but contingent on the surrounding intervention landscape, a contingency that elasticity decomposition of $R_e(t)$ makes explicit and quantifiable.
