📄 infectious diseases

Optimal vaccination in aging populations under age-dependent infection fatality risks

This study uses transmission models to demonstrate that in aging populations, the optimal timing and frequency of vaccination depend critically on pathogen transmissibility and immunity duration, as strategies that suppress infections early without providing long-term protection can paradoxically shift infections to older ages with higher fatality risks, thereby increasing overall mortality and years of life lost.

Original authors: van Boven, M., van Dorp, C., Bosschaert, M., van der Schans, J., van Baarle, D., Kretzschmar, M. E.

Published 2026-06-26

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: van Boven, M., van Dorp, C., Bosschaert, M., van der Schans, J., van Baarle, D., Kretzschmar, M. E.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

The Big Picture: The "Aging Crowd" Problem

Imagine a town where the population is getting older and older. In this town, a nasty flu-like virus circulates. The scary part is that this virus is like a "steep hill": it's very easy to survive if you are young, but if you fall on the hill when you are very old, the fall is much more dangerous.

The researchers asked a simple question: When should we give out the "shield" (the vaccine) to the people in this town to save the most lives?

They used a computer simulation (a digital model of the town) to test different rules. They wanted to see if giving the shield to kids, middle-aged adults, or the elderly worked best, and whether giving it out once or every year made a difference.

The Two Main Scenarios

The paper looks at two very different ways of using the vaccine, and they lead to opposite results.

1. The "One-Time Shield" (Long-Lasting Protection)

Imagine a vaccine that acts like a super-strong, permanent shield that lasts for decades (like 10 to 50 years).

The Finding: If the virus is very contagious, the best time to give this shield is actually when people are older (around 60 to 80 years old).
The Analogy: Think of it like putting a roof on a house. If you put a roof on a house that is already old and fragile, you save the most damage. If you put a roof on a brand-new house (a child) that is already sturdy, you aren't saving much extra life because the house was going to be fine anyway.
Why? Since the shield lasts a long time, you don't need to give it to kids to protect them later. You save the most lives by protecting the people right before they reach the "danger zone" of the steep hill.

2. The "Annual Umbrella" (Short-Lasting Protection)

Now, imagine a vaccine that is like a cheap, flimsy umbrella that only works for a few months or a year. This is more like how real flu vaccines work today. You have to get a new one every year.

The Finding: This is where it gets tricky. If the virus is very contagious and the umbrella only lasts a short time, giving the umbrella to young people can actually make things worse.
The Analogy: Imagine a game of "Musical Chairs" where the chairs are "safe spots."
- If you give the umbrella to young people, they stay safe for a while.
- But because the umbrella wears out quickly, they eventually lose their protection.
- Meanwhile, the virus is still out there. Because the young people were protected for a while, they didn't catch the virus early.
- The Trap: Now, the virus is waiting for everyone to get older. When the young people finally lose their protection, they are now older. They catch the virus at an age where the "fall" is much more dangerous.
- The Result: The vaccine didn't stop the virus; it just pushed the infection to an older age. Since the virus is deadlier for older people, the total number of deaths actually goes up compared to having no vaccine at all.

The "Age-Shifting" Danger

The paper calls this the "Age-Shifting" effect.

Think of it like a traffic jam. If you put a speed bump (vaccine) in the middle of the road, cars (viruses) might slow down and stop there. But if the speed bump disappears quickly, the cars don't disappear; they just keep driving until they hit a cliff (old age) where the crash is fatal.

The study found that if you vaccinate young people with a short-lived vaccine against a highly contagious virus, you are essentially delaying the infection until the person is too old to survive it as easily.

Key Takeaways from the Study

There is no "One Size Fits All" rule. You can't just say "vaccinate everyone at age 50." It depends entirely on how contagious the virus is and how long the vaccine lasts.
Long-lasting vaccines: Best given later in life (60s–80s) to protect people right before they get frail.
Short-lasting vaccines (like the flu shot): Can be dangerous if given too early to young people when the virus is very contagious. It might push the infection to the "danger zone" of old age, increasing deaths.
The "Perfect" Strategy: To avoid the "Age-Shifting" trap, you need to balance the timing. Sometimes, it's better to wait until people are older to start the annual vaccination campaign, so you don't accidentally push the infection into the most vulnerable years.

What the Study Doesn't Say

The authors are careful to note that their model assumes the vaccine is an "all-or-nothing" shield (it either works perfectly or not at all). In the real world, vaccines often just make the illness milder, even if they don't stop the infection completely. This might mean the "Age-Shifting" danger isn't quite as bad in real life as the model predicts, but the principle remains: we must be careful about when we vaccinate to avoid accidentally making the problem worse for the elderly.

In short: Vaccination is a powerful tool, but if you use a short-term shield against a fast-moving virus, you might accidentally push the danger onto the shoulders of the oldest and most vulnerable people.

Problem Statement
As global populations age, the burden of respiratory infections such as influenza A, respiratory syncytial virus (RSV), and SARS-CoV-2 is shifting toward older adults. A critical characteristic of these pathogens is that infection fatality rates (IFR) rise steeply with age. While vaccination programs have successfully reduced disease burdens in childhood, the optimal timing and frequency of adult vaccination remain unclear. There is a specific concern that vaccination strategies which suppress infections in younger or middle-aged groups without providing sustained protection into late life may inadvertently shift the age distribution of infections toward older cohorts where fatality risks are highest. This "age-shifting" effect could potentially increase overall mortality and years of life lost (YLL), particularly in the context of waning immunity and demographic transitions.

Methodology
The authors employed age-structured transmission models (SIRV) incorporating demographic projections, age-specific IFRs, empirical contact patterns, and waning immunity. The study utilized a demographic baseline based on mortality projections for the Netherlands, approximated by a log-linear function, with a fixed annual birth cohort and zero migration to maintain generality.

Key model components included:

Transmission Dynamics: A time- and age-dependent model where individuals transition between Susceptible (S), Infected (I), naturally Immune (R), and Vaccinated (V) compartments. Immunity durations were modeled using Erlang distributions to allow for realistic variability.
Vaccination Scenarios: Two primary scenarios were analyzed:
1. Single Vaccination: A one-time dose administered at a specific age ( $a_v$ ) with long-lived protection (5–50 years).
2. Repeated Vaccination: Annual vaccinations (mimicking influenza campaigns) starting at a specific age, with short-lived vaccine-induced immunity (0.5–5 years) compared to natural immunity (2–7 years).
Parameters: The study varied the basic reproduction number ( $R_0$ ), vaccination coverage (80% and 30% effective), age at first vaccination, and the duration of immunity.
Outcomes: The models calculated total infection-induced mortality and Years of Life Lost (YLL) over a long-term horizon (50–150 years) to identify vaccination schedules that minimize these burdens.

Key Results
The study found that the optimal vaccination strategy is highly dependent on the interplay between pathogen transmissibility, the duration of immunity, and population demography.

Single Vaccination with Long-Lived Protection: For pathogens with strongly age-increasing fatality risks, the age minimizing mortality in older adults typically ranges from 60 to 80 years. This optimal age shifts toward older ages when transmissibility ( $R_0$ ) is higher or natural immunity lasts longer. However, for low $R_0$ and long-lasting immunity, vaccinating young children remains optimal due to herd effects. In cases of high $R_0$ combined with long-lasting immunity, single-dose vaccination can become counterproductive, increasing mortality by postponing infections to older, more vulnerable ages.
Repeated Vaccination with Short-Lived Immunity: This scenario produced qualitatively different and potentially adverse outcomes. When vaccine-induced immunity is short-lived (<5 years) and transmissibility is high, initiating vaccination at younger ages can increase total mortality and YLL compared to no vaccination. This occurs because vaccination suppresses or delays infections in midlife but fails to protect individuals into late life, effectively shifting the infection burden to the oldest age groups where fatality risks are exponential.
Coverage Sensitivity: Lower effective vaccination coverage (e.g., 30%) attenuated mortality reductions but expanded the parameter space where repeated vaccination increases mortality, particularly when vaccine protection is short-lived relative to natural immunity.
Demographic Influence: Changes in background mortality schedules (comparing 2025 vs. 2050 projections) had only a modest influence on the optimal vaccination age (shifting it by 1–3 years). The results were driven primarily by the steepness of the age gradient in IFR rather than the absolute scale of mortality.

Significance and Claims
The authors conclude that there is no universal rule for the optimal timing of adult vaccination. The findings highlight that vaccination strategies must be evaluated across the full life course, considering how they alter the age distribution of infections over decades.

The paper claims that:

Optimal timing is context-dependent: It relies jointly on pathogen transmissibility, the duration of natural and vaccine-induced immunity, and demographic structure.
Risk of unintended consequences: Vaccination programs that suppress infections earlier in life without ensuring protection into late life may shift infections toward ages with higher fatality risks, potentially increasing the long-term disease burden.
Policy Implications: These findings provide a cautionary benchmark for designing vaccination strategies against influenza A and other pathogens with strongly age-dependent IFRs. As vaccine formulations evolve (potentially offering stronger but shorter protection), policymakers must account for life-history effects and the redistribution of infection risks, rather than focusing solely on immediate, per-season protection.

The authors acknowledge limitations, noting that their models assume "all-or-nothing" vaccines that fully prevent infection, which may overestimate age-shifting effects compared to real-world vaccines that primarily reduce severity. Additionally, the models assume monotonically increasing IFRs, which may not apply to pathogens with U-shaped severity profiles (e.g., RSV in infants). Despite these simplifications, the study serves as a foundational step in understanding the long-term population-level impacts of adult immunization.