Many respiratory viruses follow a seasonal cycle characterized by a single annual infection wave. Pandemic respiratory viruses, however, frequently deviate from this pattern and generate multiple waves of infection over short intervals. Among several hypothesized mechanistic drivers of these multi‑wave dynamics, the emergence of new variants carrying immune‑escape mutations is prominent: such variants can infect individuals who were previously immune to earlier virus strains.
The paper introduces a new modeling approach intended to capture these biologically grounded mechanisms within an epidemiological framework rather than relying on retrofitted parameter changes during ongoing outbreaks.
Existing epidemic models derived from the Susceptible‑Infectious‑Recovered (SIR) family and their common extensions provide foundational insight into transmission dynamics but do not explicitly model variant emergence or the evolving immune landscape of the host population. As a consequence, practitioners commonly perform ad hoc parameter adjustments during simulations to reproduce observed multi‑wave case counts. These adjustments reduce explanatory clarity and undermine mechanistic interpretation of drivers behind successive waves.
To address those gaps, the authors developed the Immunity‑Variants‑Epidemic (IV‑Epidemic) model. The IV‑Epidemic framework integrates candidate biological and epidemiological drivers of multi‑wave epidemics — notably variant evolution with immune escape and a dynamic population immune profile — into a single mathematical model. The structure of the model is designed to allow new, genetically distinct variants to arise and to interact with pre‑existing immunity in the host population, producing infection waves without manual in‑simulation parameter tuning.
The authors present IV‑Epidemic as a unifying approach that explicitly represents how antigenic change and population immunity jointly shape epidemic trajectories.
Model parameters were calibrated using publicly available SARS‑CoV‑2 data. Reported case counts and related epidemiological data were obtained from Our World in Data. State‑level policy stringency inputs were taken from the Oxford COVID‑19 Government Response Tracker (OxCGRT). The manuscript states that only aggregate, openly accessible human data were used, and that no restricted or identifiable individual‑level data contributed to the analysis.
Specific numerical parameter values, model equations, and calibration procedures are presented in the full preprint; this summary reflects the high‑level approach and the data sources used for fitting.
After calibration, the IV‑Epidemic model closely reproduced observed multi‑wave infection trajectories seen in the SARS‑CoV‑2 pandemic. Importantly, the model achieved realistic multi‑wave patterns using primitive model inputs and without dynamic in‑simulation parameter modifications, demonstrating that mechanisms represented in the model can generate multi‑wave behavior endogenously.
This outcome contrasts with standard SIR‑type simulations that often require external parameter changes to mimic repeated waves.
Beyond aggregate case counts, the IV‑Epidemic simulations generated distributions of infections across concurrently circulating variants. These simulated distributions align with observed epidemic behavior in which individual waves are typically driven by a relatively small number of emerging, genetically distinct variants. The model therefore offers a mechanistic account of how variant emergence and competitive dynamics among variants contribute to the timing and magnitude of successive infection waves.
The IV‑Epidemic results underscore the substantial influence of pre‑existing immunity on early epidemic spread. The model also highlights how the evolving immune profile of the population — shaped by prior infections, variant‑specific immune escape, and presumably interventions — can drive subsequent infection patterns. By representing immunity as a population‑level, time‑varying factor that interacts with variant properties, IV‑Epidemic provides a framework to examine how immunity landscapes alter susceptibility and transmission over the course of a pandemic.
By explicitly incorporating variant evolution and population immunity, the IV‑Epidemic model is proposed as a tool to strengthen the predictive and explanatory power of epidemiological surveillance systems. Because it can reproduce multi‑wave dynamics and variant distributions from mechanistic inputs, IV‑Epidemic could support more interpretable scenario analysis and help disentangle the roles of viral evolution and host immunity during pandemics.
The manuscript frames the model as a potential complement to surveillance and modeling toolkits, with the goal of improving mechanistic understanding rather than offering immediate clinical guidance.
The authors report that the study used only openly available, aggregate data (Our World in Data for COVID‑19 case counts; OxCGRT for policy stringency). All data produced in the study are available upon reasonable request to the authors. The preprint notes that ethical guidelines were followed, no competing interests were declared, and that the work has not been peer reviewed. The authors caution that the preprint should not be used to guide clinical practice.