Abstract
OBJECTIVE: This study aimed to investigate the spatiotemporal heterogeneity of human brucellosis and quantify the exposure-lag-response relationships of key socioeconomic and livestock production drivers in Ningxia, China, from 2007 to 2022. The goal was to generate evidence for developing targeted, integrated interventions in this high-burden pastoral region. METHODS: We conducted a retrospective ecological study integrating human brucellosis surveillance data with county-level socioeconomic and livestock production statistics. A multi-analytic framework was employed: Joinpoint regression analyzed long-term trends; spatiotemporal scan statistics identified high-risk clusters; GeoDetector quantified the explanatory power of potential drivers on spatial heterogeneity; and Distributed Lag Nonlinear Models (DLNMs) were constructed to assess the nonlinear and lagged effects of significant drivers on monthly incidence. RESULTS: The human brucellosis incidence rate in Ningxia increased 167-fold, from 0.52 to 86.83 per 100,000 population between 2007 and 2022. Spatiotemporal analysis revealed a persistent high-risk cluster (Relative Risk, RR = 4.22, P < 0.001) in 11 eastern counties. GeoDetector identified livestock-related factors as primary spatial drivers, with sheep inventory (q = 0.96) and cattle inventory (q = 0.92) showing the highest explanatory power. DLNM results indicated a significant 3-year lagged risk associated with low cattle stocking levels (RR = 2.75), while sheep stocking exhibited a complex, non-linear U-shaped lag effect. In contrast, higher regional Gross Domestic Product (GDP) was associated with an immediate lower risk (RR = 0.81). CONCLUSION: The brucellosis epidemic in Ningxia is characterized by intense spatial clustering and is associated with distinct, lagged effects of livestock production structures coupled with immediate economic influences. The findings underscore that livestock production metrics can serve as effective proxies for risk mapping even in the absence of direct animal infection data. Our study highlights the necessity for a dual-strategy intervention: implementing risk-based veterinary public health measures in high-incidence clusters while leveraging economic development to strengthen long-term prevention and control capacities.