Incoprea Theory
Volume VI · 2026

Theory 11/Public Health & Epidemiology /Vaccine epidemiology/3 min

Effectiveness Estimates Inherit History

The same vaccine, two cohorts, two numbers. The difference may be in what their immune systems had already seen.

The claim

Vaccine effectiveness estimates are biased by prior exposure history rather than only by confounding, so cohorts with different infection histories should yield different effectiveness for the same vaccine, and the bias should be predictable from serology.

01The evidence

  • Existing immunity has been shown to bias or limit current responses, particularly against evolving viruses, which means prior exposure changes the response a vaccine can produce.Immunity
  • Observational effectiveness studies have reported values lower than expected, and the discrepancy is typically attributed to confounding or waning rather than to the composition of prior immunity in the cohort.Clinical Infectious Diseases

02The tension

Effectiveness is treated as a property of the vaccine, estimated in a cohort and generalised. Cohorts differ in what they have previously been infected with, and that difference changes the immune response, so the estimate is partly a property of the cohort.

03The theory

A vaccine does not produce immunity from nothing; it redirects an existing repertoire. If a cohort has been exposed to related strains, the response is shaped by that history, which can either broaden or narrow the response depending on the relationship between the strains.

If history shapes the response, then effectiveness is a function of cohort serology. Two cohorts with the same vaccine and the same strain but different histories will show different effectiveness, and the difference is not confounding in the usual sense.

The prediction is that effectiveness should vary between cohorts in a way predicted by pre-vaccination serology, and that adjusting for serology should reduce the between-study heterogeneity that is currently attributed to method.

The practical consequence is that effectiveness estimates should carry their cohort's history as a caveat, and that pooling across cohorts without serology assumes an exchangeability that does not hold.

04What this is not

Confounding-and-waning explanations

Those explanations treat between-study variation as methodological or temporal. This treats it as compositional, which predicts variation from serology measured before vaccination.

05Predictions

  • P1Effectiveness should vary between cohorts in proportion to prior exposure history.
  • P2Adjusting for serology should reduce between-study heterogeneity.
  • P3Cohorts with narrow prior exposure should show different effectiveness from broadly exposed ones.

06What would refute it

Effectiveness constant across cohorts at controlled serology would leave confounding as the explanation.

07What it would change

It would add serology to effectiveness reporting and change how estimates are pooled.

08Evidence base

  1. Existing immunity biases current responses to evolving viruses — Immunityhttps://www.cell.com/action/showPdf?pii=S1074-7613%2824%2900315-7
  2. Observational vaccine effectiveness lower than expected — Clinical Infectious Diseaseshttps://academic.oup.com/cid/advance-article-pdf/doi/10.1093/cid/ciae132/57233798/ciae132.pdf