A time-space Bayesian regression model of rabies cases in the animal population of Kazakhstan (2013–2023)

Abstract

Introduction: Despite its endemic status and socioeconomic impacts, the spatial-temporal variation in rabies risk and its underlying determinants in Kazakhstan animal populations remain poorly understood. This study aimed to characterize the time-space dynamics of rabies in animal populations across Kazakhstan regions from 2013 to 2023 and identify the key drivers of transmission. Methods:Using a Bayesian hierarchical regression model with spatial and temporal random effects, we analyzed national surveillance data on rabies cases in livestock, companion animals, and wildlife, alongside sociodemographic and animal population variables. Results:The model revealed that higher median income (odds ratio [OR]: 1.18, 95% posterior predictive interval [PPI]: 1.06–1.31), the presence of rabies in wildlife (OR: 1.55, 95% PPI: 1.27–1.89), and companion animal rabies incidence (low: 1–5 cases/year, OR: 1.39, 95% PPI: 1.06–1.85; high: ≥6 cases/year, OR: 2.07, 95% PPI: 1.46–2.96) were associated with increased livestock rabies risk, while higher human population density correlated with reduced risk (OR: 0.68, 95% PPI: 0.5–0.9). Spatial analysis identified persistent high-risk zones in western Kazakhstan and lower risk in southern regions, driven by ecological and socioeconomic heterogeneity. Discussion:These findings highlight the relationship between wildlife reservoirs, domestic animal management, and socioeconomic factors in rabies transmission in Kazakhstan. By integrating these insights into national policy, Kazakhstan can advance toward the global target of eliminating dog-mediated human rabies deaths by 2030, serving as a model for Central Asia.