Race-specific estimates of the force of infection of Helicobacter pylori using two national seroprevalence surveys in the US: A hierarchical Bayesian catalytic epidemic modeling study

Preprint — SSRN

Forthcoming
Hierarchical Bayesian catalytic modelling of race-specific Helicobacter pylori force of infection using two national seroprevalence surveys in the United States. Preprint on SSRN.
Authors

Roa, J.

Peralta, Y.

McMahon, M.

Taylor, C.

Ward, Z.J.

Camargo, M.C.

Yeh, J.

Alarid-Escudero, F.

Published

January 1, 2025

Race-Specific Force of Infection of H. pylori in the US

Roa, J., Peralta, Y., McMahon, M., Taylor, C., Ward, Z.J., Camargo, M.C., Yeh, J., & Alarid-Escudero, F.

Edward Hopper — Nighthawks (1942) Art Institute of Chicago · Chicago A late-night diner glows against an empty street — four strangers sharing a space without sharing a story. Hopper painted solitude as something structural, built into the city itself, legible in the gap between people who live side by side but rarely touch.

The Problem

Helicobacter pylori is the primary risk factor for gastric cancer, and its prevalence in the United States varies dramatically across racial and ethnic groups. Yet we lack estimates of the force of infection – the rate at which susceptible individuals become infected – stratified by race, ethnicity, birthplace, and birth cohort. Without these estimates, public health interventions cannot be targeted where they are needed most.

The Approach

We developed a hierarchical Bayesian catalytic epidemic model using seroprevalence data from two nationally representative NHANES cycles (1988–1994 and 1999–2000). The model estimated age-specific prevalence and force of infection for four racial and ethnic groups – Hispanic, non-Hispanic Black, non-Hispanic White, and other races – while incorporating sex, birth cohort, and place of birth as covariates. The hierarchical structure pooled information across groups, yielding stable subgroup-specific estimates even for smaller populations.

What We Found

25,000
Individuals across two NHANES cycles
90%
Max prevalence in foreign-born populations
Childhood
Peak FOI before age 10, then rapid decline

The force of infection peaked sharply in early childhood (before age 10) and declined rapidly thereafter across all groups. Hispanic and non-Hispanic Black populations had the highest FOI peaks (0.04–0.05 infections/person/year), roughly double the rate among non-Hispanic Whites (0.02–0.03). Foreign-born individuals reached maximum prevalence near 90% by age 90, compared to 40–70% among US-born individuals. Between NHANES cycles, FOI decreased – particularly among Hispanic and non-Hispanic Black populations – indicating reduced transmission in more recent cohorts. No significant differences by sex were observed.

Estimated H. pylori seroprevalence by age, race, and NHANES cycle from a Bayesian catalytic model Estimated weighted seroprevalence of H. pylori by age, race, and NHANES cycle (Bayesian catalytic model). NH = non-Hispanic.

Estimated H. pylori seroprevalence and force of infection by NHANES cycle, peaking in early childhood
Estimated weighted seroprevalence (top) and force of infection in new infections per susceptible individual per year (bottom) of H. pylori by NHANES cycle. The FOI peaks sharply in early childhood and declines rapidly, with Hispanic and NH Black populations showing the highest infection rates.

Why It Matters

Early childhood is the critical window for H. pylori transmission, and the force of infection is substantially higher among Hispanic and non-Hispanic Black populations. These race-specific estimates provide the epidemiological foundation for targeted prevention strategies – particularly among foreign-born and minority communities that carry the greatest burden of infection and, ultimately, gastric cancer.

Citation

Roa, J., Peralta, Y., McMahon, M., Taylor, C., Ward, Z.J., Camargo, M.C., Yeh, J., & Alarid-Escudero, F. (2025). Race-specific estimates of the force of infection of Helicobacter pylori using two national seroprevalence surveys in the US: A hierarchical Bayesian catalytic epidemic modeling study. SSRN Preprint. Available at: https://ssrn.com/abstract=5677467

Citation

BibTeX citation:
@misc{j.2025,
  author = {J. , Roa and Y. , Peralta and M. , McMahon and C. , Taylor
    and Z.J. , Ward and M.C. , Camargo and J. , Yeh and F. ,
    Alarid-Escudero},
  title = {Race-Specific Estimates of the Force of Infection of
    {Helicobacter} Pylori Using Two National Seroprevalence Surveys in
    the {US:} {A} Hierarchical {Bayesian} Catalytic Epidemic Modeling
    Study},
  date = {2025-01-01},
  url = {https://jorgeroac.com/publications/papers/forthcoming/hpylori-force-infection/},
  langid = {en}
}
For attribution, please cite this work as:
J., Roa, Peralta Y., McMahon M., et al. 2025. “Race-Specific Estimates of the Force of Infection of Helicobacter Pylori Using Two National Seroprevalence Surveys in the US: A Hierarchical Bayesian Catalytic Epidemic Modeling Study.” In SSRN. Preprint, January 1. https://jorgeroac.com/publications/papers/forthcoming/hpylori-force-infection/.