Herd Immunity Threshold

Herd immunity threshold (HIT, denoted p_c or q_c) is the minimum critical proportion of a host population that must possess sterilizing or transmission-blocking immunity—acquired primarily through mass vaccination or prior infection—so that the effective reproduction number (R_e or R_t) falls below 1.0, causing ongoing chains of pathogen transmission to decay and protecting remaining susceptible individuals indirectly. Within mathematical epidemiology, biostatistics, and public health, the herd immunity threshold bridges individual immunology (neutralizing antibody titers and mucosal T-cell memory) with population network dynamics. By reducing the spatial and topological density of susceptible hosts, maintaining immunization coverage above p_c shields neonates too young for vaccination (reducing infant mortality), pregnant women, and immunocompromised patients undergoing chemotherapy or organ transplantation.

The phenomenon of indirect population protection was first observed experimentally in veterinary bacteriology by Adolph Eichhorn and George Potter (1916) and formalized quantitatively by W. W. C. Topley and G. S. Wilson (1923) in murine Salmonella epidemics, Sheldon Dudley (1924) in human diphtheria outbreaks, and A. W. Hedrich (1933) in Baltimore measles cycles. Its modern mathematical synthesis with compartmental epidemic theory was established by C. E. Gordon Smith (1970), Paul Fine (1993), and Roy M. Anderson and Robert M. May (1985, 1991).

Basic Reproduction Number (R₀), Effective Reproduction Number (R_e), and p_c = 1 − 1/R₀

In the classical Kermack–McKendrick (1927) compartmental SIR model for a homogeneously mixing closed population of size N = S + I + R, the rate of change of the infectious compartment I(t) is governed by:

dI / dt = ( β S I / N ) − γ I = γ I [ (β / γ) (S / N) − 1 ]

where β is the effective transmission rate per unit time (the product of the per-capita contact rate c and the per-contact transmission probability τ) and γ is the per-capita recovery or removal rate (with 1/γ equal to the mean infectious period D).

In a naive population where nearly all individuals are susceptible (S ≈ N), a single primary infectious case generates an expected number of secondary cases defined as the Basic Reproduction Number (R₀):

R₀ = β / γ = c × τ × D

When a proportion p of the population is completely immune to infection, the remaining susceptible fraction of the population is s = S / N = 1 − p. Because each contact made by an infectious individual now has only a probability (1 − p) of encountering a susceptible host, the expected number of secondary infections generated per primary case is reduced to the Effective Reproduction Number (R_e):

R_e = R₀ × (S / N) = R₀ (1 − p)

Substituting R_e into the infectious differential equation yields:

dI / dt = γ I ( R_e − 1 )

Epidemic incidence grows (dI/dt > 0) whenever R_e > 1, remains endemic in steady state (dI/dt = 0) when R_e = 1, and decays exponentially toward extinction (dI/dt < 0) whenever R_e < 1. Setting the threshold condition R_e = 1 and solving for the immune fraction p yields the Herd Immunity Threshold (p_c):

1 = R₀ (1 − p_c) ⇒ 1 − p_c = 1 / R₀ ⇒ p_c = 1 − (1 / R₀)

Equivalently, the critical maximum fraction of the population that can remain susceptible without permitting sustained epidemic amplification is s_c = 1 / R₀.

All-or-Nothing and Leaky Vaccine Efficacy Against Transmission (E)

In practice, no vaccine confers 100% sterilizing protection in every recipient, nor is vaccine potency preserved if the cold chain is compromised. If a vaccine exhibits a net vaccine efficacy against infection and onward transmission of E (where 0 < E ≤ 1), then vaccinating a proportion V of the population generates an effectively immune proportion p = V × E.

Substituting p = V_c × E into the herd immunity equation p_c = 1 − (1 / R₀) yields the Critical Vaccination Coverage (V_c) required to interrupt endemic circulation:

V_c = p_c / E = [ 1 − (1 / R₀) ] / E

This equation reveals a critical boundary condition in vaccinology: if vaccine efficacy against transmission E is smaller than the herd immunity threshold p_c = 1 − (1 / R₀), then even vaccinating 100% of the entire human population (V = 1.0) cannot achieve herd immunity (V_c > 1.0). For example, if a pathogen has R₀ = 5.0 (so p_c = 1 − 1/5 = 0.80, or 80%), a vaccine with E = 0.90 requires vaccinating V_c = 0.80 / 0.90 = 88.9% of the population; however, if vaccine efficacy against transmission drops to E = 0.70, then V_c = 0.80 / 0.70 = 114.3%, meaning herd immunity is mathematically unattainable via vaccination alone without non-pharmaceutical interventions that lower R₀.

Contact Network Heterogeneity and the Next-Generation Matrix

The standard formula p_c = 1 − 1/R₀ assumes homogeneous random mixing (mass-action kinetics), in which every individual in the population has an identical contact rate c. In real human societies, interpersonal contact networks exhibit high degree variance (σ_c²) and superspreading (quantified by Lloyd-Smith et al., 2005, via the negative binomial dispersion parameter k < 0.2, where 10% to 20% of infected hosts generate 80% of secondary transmissions).

As shown by Roy Anderson and Robert May (1991), in a heterogeneous network with mean contact rate μ_c and variance σ_c², the basic reproduction number is amplified by the square of the coefficient of variation (CV = σ_c / μ_c):

R₀ = R_0,hom × ( 1 + CV² ) = ( τ D μ_c ) × [ 1 + (σ_c² / μ_c²) ]

Heterogeneity alters the herd immunity threshold depending on how immunity is acquired:

1. Uniform Random Vaccination: Remains governed by the higher heterogeneous R₀, unless vaccination is specifically targeted to high-contact hub nodes (such as healthcare workers, transit operators, and dense urban cohorts), which lowers the required population coverage V_c substantially below 1 − 1/R₀.
2. Assortative Clustering of Unvaccinated Subpopulations ("Homophily"): A dangerous departure from homogeneous mixing occurs when unvaccinated individuals are not distributed randomly across society, but instead cluster spatially or socially in schools, neighborhoods, or communities with high internal contact rates (assortative mixing). Represented mathematically by the Next-Generation Matrix (NGM, K) with elements k_ij = R_0,*ij* s_i (the expected secondary infections in subgroup i caused by an infected individual in subgroup j), epidemic persistence is governed by the dominant eigenvalue (spectral radius ρ(K)) of the matrix. Even when a nation's overall mean vaccination coverage exceeds 95%—well above the nominal 93% herd immunity threshold for measles—a localized neighborhood or school where coverage drops to 65% forms a sub-network with local spectral radius R_e,local > 1.0, sustaining explosive outbreak clusters and threatening health equity.

Waning Mucosal Immunity, Antigenic Evolution, and Non-Human Zoonotic Reservoirs

Classical sterilizing herd immunity—culminating in the permanent global eradication of smallpox (Variola virus, certified in 1980) and the veterinary morbillivirus rinderpest (certified in 2011)—requires four biological conditions that not all pathogens satisfy:

• Lifelong Sterilizing Immunity vs. Waning Mucosal IgA: Systemic viral infections with an obligate viremic phase and long incubation period (such as measles, rubella, and smallpox) allow memory B cells time to mount an anamnestic IgG response that clears the virus before secondary shedding occurs. In contrast, respiratory viruses that replicate rapidly on the superficial nasal and bronchial epithelium with a 1- to 3-day incubation period (such as SARS-CoV-2, respiratory syncytial virus, and rhinoviruses) rely on transient secretory mucosal IgA antibodies that wane with a half-life of months (SIRS compartments), permitting reinfection even when systemic memory T cells continue to protect against severe hospitalization and years of potential life lost.
• Antigenic Drift and Escape Mutations: RNA viruses replicated by error-prone RNA polymerase accumulate surface glycoprotein mutations (such as influenza hemagglutinin drift or coronavirus spike receptor-binding domain substitutions) that erode the neutralizing efficacy E of prior immunity, continuously shifting the herd immunity target until updated multivalent vaccines are deployed and monitored via wastewater-based epidemiology.
• Absence of Non-Human Wildlife or Environmental Reservoirs: True eradication via herd immunity is only possible for anthroponotic pathogens restricted exclusively to human hosts (such as smallpox, measles, and poliovirus). Pathogens maintained in sylvatic wildlife or livestock reservoirs across the One Health interface (such as yellow fever in non-human primates, rabies in chiropteran/canid hosts, or avian influenza in waterfowl) or environmental spore formers (Clostridium tetani in soil, for which the tetanus toxoid vaccine protects the individual against tetanospasmin neurotoxin but provides zero herd immunity because tetanus is not transmitted person-to-person) require continuous lifelong immunization of every new birth cohort.

1. Fine, P. E. M. (1993). "Herd immunity: history, theory, practice." Epidemiologic Reviews, 15(2), 265–302. https://doi.org/10.1093/oxfordjournals.epirev.a036121
2. Anderson, R. M., & May, R. M. (1985). "Vaccination and herd immunity to infectious diseases." Nature, 318(6044), 323–329. https://doi.org/10.1038/318323a0
3. Fine, P., Eames, K., & Heymann, D. L. (2011). "'Herd immunity': a rough guide." Clinical Infectious Diseases, 52(7), 911–916. https://doi.org/10.1093/cid/cir007
4. Smith, C. E. G. (1970). "Prospects for the control of infectious disease." Proceedings of the Royal Society of Medicine, 63(11 Part 2), 1181–1190. https://doi.org/10.1177/003591577006311P206
5. Lloyd-Smith, J. O., Schreiber, S. J., Kopp, P. E., & Getz, W. M. (2005). "Superspreading and the effect of individual variation on disease emergence." Nature, 438(7066), 355–359. https://doi.org/10.1038/nature04153