Classical mean-field epidemic theory upper-bounds real disease spread

Journal Article (2026)
Author(s)

Brian L. Chang (TU Delft - Electrical Engineering, Mathematics and Computer Science)

Piet Van Mieghem (TU Delft - Electrical Engineering, Mathematics and Computer Science)

Research Group
Network Architectures and Services
DOI related publication
https://doi.org/10.1038/s42005-026-02723-3 Final published version
More Info
expand_more
Publication Year
2026
Language
English
Research Group
Network Architectures and Services
Journal title
Communications Physics
Issue number
1
Volume number
9
Article number
320
Page Views
4
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

Non-Markovian theory provides the most accurate stochastic description of an infection spreading on a contact graph with a general distribution of the infection time and curing time. However, the infection time is not directly measurable and instead the generation time is measured in real-world epidemics. Here we show that if the generation time follows approximately a Weibull distribution with shape parameter α, then the Weibull infection time of the spreading process possesses a similar shape parameter α. For a large list of published generation times in biological, mainly respiratory epidemics, the Weibull shape parameter α is computed, which reveals that nearly always α > 1. The impact of this observation is that a complicated non-Markovian analysis is not needed, because the well-established Markovian case α = 1 provides a lower bound for the epidemic threshold and an upper bound for the basic reproduction number and the nodal infection probabilities in the contact graph.