📄 medicine

The Cross-Immunity Valley: Non-Monotonic Dynamics and the Hospital-Morbidity Trade-off in Delayed Variant Emergence

This study reveals a fundamental epidemiological trade-off in which delaying the introduction of an immune-evasive variant during an ongoing epidemic creates a "Cross-Immunity Valley" that minimizes acute hospital peaks but maximizes long-term cumulative infections, whereas simultaneous emergence suppresses long-term burden at the cost of a catastrophic acute peak, implying that the efficacy of travel restrictions depends critically on the variant's immune-evasion phenotype and the timing of its arrival.

Original authors: Arjun Garg

Published 2026-08-27

📖 7 min read🧠 Deep dive

CC BY 4.0

Original authors: Arjun Garg

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

When a new virus strain emerges, public health officials face a difficult choice: do they try to stop it from entering a country, or do they let it in and hope the population is already protected? This question sits at the intersection of epidemiology, the study of how diseases spread, and immunology, the study of how the body defends itself. The core idea is that when a virus infects a person, the body builds a defense that usually protects against that specific virus and sometimes against similar ones. However, viruses change. A new variant might look different enough to slip past these defenses, a trait scientists call immune evasion. The timing of when this new variant arrives matters immensely. If it arrives while the first virus is still sweeping through a population, many people are either currently sick or have just recovered and are still protected. If it arrives months later, that protection may have faded, leaving the population vulnerable again. Understanding how these factors interact is crucial for deciding whether travel bans and border closures actually save lives or simply shift the burden of disease to a later time.

A researcher named Arjun Garg, working at a high school for science and technology, used a computer model to explore this exact dilemma. The study focused on a scenario where a new, slightly more contagious virus variant appears while an older version is still circulating. The model simulated a population of one million people and tracked how the two versions of the virus competed over a period of 900 days. The researcher varied two main factors: how long it took for the new variant to arrive after the first one started spreading, and how well the new variant could bypass the immunity built up by the first one. The goal was to see how these choices affected two specific outcomes: the highest number of people sick at the same time, which determines the strain on hospitals, and the total number of people who got sick over the entire 900-day period.

The simulations revealed a surprising and counterintuitive pattern. When the new variant arrived at the exact same time as the old one, the two strains fought for the same group of susceptible people immediately. Because the new variant was slightly better at spreading, it quickly pushed the old one out of the picture. This resulted in a single, massive wave of infection that overwhelmed hospitals with a peak of about 344,000 people sick at once. However, because the competition happened all at once, the total number of people infected over the long term was the lowest of all scenarios, totaling roughly 2.36 million cases. The system resolved itself quickly because the stronger strain took over completely.

The story changed dramatically when the arrival of the new variant was delayed. If the variant arrived about 30 to 100 days after the first wave began, it entered a population where a large number of people were still recovering from the first virus. These recently recovered individuals possessed a form of cross-protection that made it harder for the new variant to infect them. This created a natural barrier, slowing the new virus down and flattening its peak. In these simulations, the peak number of sick people dropped by more than half, to around 160,000. This phenomenon, which the author calls the "Cross-Immunity Valley," acts as a shield for hospitals, keeping the immediate pressure on healthcare systems manageable.

However, this safety for hospitals came with a hidden cost. By delaying the new variant, the two viruses did not fight each other immediately. Instead, the first virus finished its cycle, and then the second virus started its own separate cycle later. This meant the population suffered through two distinct waves of illness rather than one. As a result, the total number of people infected over the 900 days rose to its highest point, reaching approximately 2.81 million. The delay that saved the hospitals in the short term ended up infecting more people in the long run. The study suggests that the strategy of buying time by delaying the variant's entry is a trade-off: it protects the immediate capacity of hospitals but increases the total number of infections over time.

This trade-off is not the same for every virus. The simulations showed that the "Cross-Immunity Valley" only exists if the new variant has a moderate ability to evade immunity. If the new variant is highly evasive, meaning it can easily infect people who were previously protected, the timing of its arrival matters much less. In these cases, the new virus behaves almost as if the population has no immunity at all, regardless of when it arrives. The protective effect of the valley disappears, and the virus causes large peaks of infection whether it arrives early or late. This implies that for highly dangerous variants, simply waiting or delaying their entry does not provide the same hospital protection that it might for less evasive strains.

The research also explored what happens if the delay is very long, stretching out to nearly a year. In this scenario, the protection from the first virus fades away completely as people's immunity wanes. The new variant then arrives to find a population that is once again largely susceptible. While this avoids the highest total infection count seen in the moderate delay scenario, it does not return to the lowest possible total. Instead, it creates a middle ground where the peak remains relatively low, but the total number of infections stays higher than if the two viruses had emerged together. This suggests that there is no single "perfect" delay time that minimizes both hospital strain and total infections simultaneously.

These findings challenge the common instinct that delaying a new virus is always the best strategy. The study indicates that the effectiveness of policies like travel restrictions depends heavily on two things: how well the new virus can dodge existing immunity and exactly when it arrives relative to the current wave of the old virus. If a new variant is not very good at evading immunity, delaying its arrival until it hits the "valley" of cross-protection is a smart move for protecting hospitals, even if it means more people will get sick in total over the following year. But if the variant is highly evasive, or if the delay is too long, that strategy loses its power. The author emphasizes that public health officials need to rapidly determine how evasive a new variant is and monitor the current state of the existing virus wave before deciding whether to close borders or restrict travel.

Ultimately, the study paints a picture of a complex balancing act. The intuitive goal of "flattening the curve" to protect hospitals can sometimes lead to a higher total burden of disease over time. The simulations show that the most effective way to reduce the total number of infections is actually to let the two viruses compete immediately, allowing the stronger one to take over quickly. While this causes a terrifyingly high peak of sickness, it resolves the epidemic faster. Conversely, spreading the infection out over time by delaying the new variant protects the healthcare system from being overwhelmed in the moment but extends the duration of the epidemic and infects more people overall. The paper concludes that these are not just abstract numbers but real policy choices, where saving a hospital bed today might mean infecting more people tomorrow, and that understanding this specific trade-off is essential for making informed decisions during a pandemic.

Technical Summary: The Cross-Immunity Valley and the Hospital-Morbidity Trade-off

Problem Statement
The sequential emergence of immune-evasive pathogen variants, exemplified by the SARS-CoV-2 pandemic, presents a critical forecasting and policy challenge. Public health authorities frequently employ delay-based strategies—such as border closures and travel restrictions—to postpone the introduction of novel variants. However, the epidemiological consequences of the timing of this introduction relative to an ongoing wildtype epidemic remain poorly quantified. Specifically, it is unclear whether delaying a variant's arrival reduces overall population harm or merely redistributes it, and how the variant's immune-evasion phenotype interacts with the population's existing immune landscape to determine acute hospital burden versus long-term cumulative morbidity.

Methodology
The study employs a deterministic, two-strain SIRS (Susceptible-Infected-Recovered-Susceptible) compartmental model to simulate epidemic dynamics in a well-mixed population of $N = 10^6$ . The model tracks five dynamic compartments: Susceptible ( $S$ ), Wildtype-infected ( $I_1$ ), Variant-infected ( $I_2$ ), Wildtype-recovered ( $R_1$ ), and Variant-recovered ( $R_2$ ).

Key features of the model formulation include:

Asymmetric Cross-Immunity: The model incorporates an immune-evasion coefficient $\eta_2 \in [0, 1]$ for the variant, representing its ability to infect individuals with wildtype immunity ( $R_1$ ). Conversely, the wildtype strain cannot evade variant immunity ( $\eta_1 = 0$ ).
Waning Immunity: Sterilizing immunity wanes at a uniform rate $\delta$ , returning individuals to full susceptibility, which drives long-term endemic cycling.
Simulation Strategy: The study systematically varies two parameters over a $50 \times 50$ $50 \times 50$ grid:
1. Introduction Delay ( $T_{intro}$ ): The time (0–500 days) after the wildtype epidemic begins before the variant is introduced.
2. Immune Evasion ( $\eta_2$ ): The variant's ability to bypass cross-protection (0.0 to 1.0).
Outputs: The simulations evaluate two primary metrics over a 900-day horizon: (i) the peak concurrent variant infections (proxy for acute hospital strain) and (ii) the total cumulative infections (proxy for long-term morbidity).

Key Contributions and Results
The research identifies a fundamental, non-monotonic trade-off between acute hospital capacity and long-term cumulative morbidity, driven by the interplay between introduction timing and cross-immunity.

The Cross-Immunity Valley:
For variants with moderate immune evasion ( $\eta_2 \lesssim 0.6$ ), a distinct "valley" exists in the parameter space where peak variant incidence is minimized. This occurs when the variant is introduced during the active phase of the wildtype wave ( $T_{intro} \approx 30–100$ days). In this window, the high prevalence of recently recovered wildtype individuals ( $R_1$ ) provides maximal cross-protective friction, suppressing the variant's effective reproduction number and flattening the acute peak by up to 53–54% compared to simultaneous emergence.
The Competitive-Exclusion Minimum:
Contrary to the intuition that delaying a variant is always beneficial for long-term outcomes, the study finds that simultaneous emergence ( $T_{intro} = 0$ ) minimizes total cumulative infections. When the fitter variant (higher $R_0$ ) emerges simultaneously with the wildtype, it rapidly competitively excludes the wildtype strain, consolidating the epidemic into a single wave. Any delay allows the wildtype to complete a more autonomous cycle, resulting in two partially independent waves and a higher total cumulative burden over 900 days.
The Trade-off:
The timing that offers the greatest protection to hospital infrastructure (the Cross-Immunity Valley) is the same timing that produces the highest total cumulative morbidity.
- Simultaneous Emergence: Lowest cumulative burden, but catastrophic acute peak.
- Valley Timing ( $T_{intro} \approx 30–100$ ): Lowest acute peak (hospital-safe), but highest cumulative burden.
- Late Emergence ( $T_{intro} > 200$ ): As immunity wanes, the "valley" effect diminishes, and cumulative burden rises again due to the loss of cross-protection, though it may not reach the peak of the valley scenario.
Role of Immune Evasion:
The existence of the Cross-Immunity Valley is contingent on the variant's immune-evasion phenotype. For highly evasive variants ( $\eta_2 \gtrsim 0.8$ ), the valley disappears entirely; cross-immunity provides insufficient friction, and peak incidence remains high regardless of introduction timing.

Significance and Claims
The paper claims to provide a formal mechanistic basis for understanding why delay-based policies can yield qualitatively different outcomes depending on the variant's immune-evasion profile and the phase of the ongoing epidemic.

Policy Implications: The efficacy of travel restrictions is not absolute but conditional. A policy designed to delay a variant into the "Cross-Immunity Valley" successfully protects hospital capacity in the short term but mathematically increases the total long-term infection burden by disrupting competitive exclusion.
Surveillance Necessity: The study argues that for delay-based policies to achieve their intended goals, rapid characterization of a variant's immune-evasion phenotype ( $\eta_2$ ) is essential. Without knowing if a variant is moderately or highly evasive, policymakers cannot determine if a delay will result in a protective valley or a futile redistribution of risk.
Reframing Success Metrics: The author suggests that the success of border restriction policies should not be measured solely by "flattening the curve" (acute peak reduction). Instead, public health communications must explicitly acknowledge the trade-off: saving hospitals in the immediate term may come at the cost of infecting a larger total proportion of the population over the medium term.

The study concludes that while the Cross-Immunity Valley is a real phenomenon for moderately evasive variants, it represents a specific temporal niche with distinct costs. The framework offers a rigorous scaffold for calibrating the trade-off between acute and chronic pandemic burdens before the next variant emerges.

Technical Summary: The Cross-Immunity Valley and the Hospital-Morbidity Trade-off

More like this