Comprehensive Dynamic Influence of Multiple Meteorological Factors on the Detection Rate of Bacterial Foodborne Diseases under Spatio-Temporal Heterogeneity

Xiaojuan Qi, Jingxian Guo, Shenjun Yao, Ting Liu, Hao Hou, Huan RenInternational Journal of Environmental Research and Public Health 2023 · DOI

TL;DR

This study investigates how meteorological factors influence the spread of Vibrio parahaemolyticus (a common cause of foodborne illness) in Zhejiang Province, China. By analyzing spatio-temporal patterns and using a Vector Autoregressive (VAR) model, researchers found that temperature, humidity, and precipitation have lagging effects on disease detection rates, with varying impacts across different regions. This highlights the need for region-specific, proactive public health interventions.

What Problem Does This Paper Address?

Foodborne diseases are a significant global health concern, leading to substantial economic losses and impacting human well-being. While it's understood that environmental factors, particularly climate, play a role in the incidence of these diseases, the precise dynamic relationships and how they vary across different geographical areas remain complex. Specifically, predicting outbreaks requires understanding not just the presence of certain weather conditions but also their delayed effects and how these effects differ between regions with varying disease clustering. This paper aims to bridge this gap by dissecting the spatio-temporal heterogeneity of bacterial foodborne diseases, focusing on Vibrio parahaemolyticus, and quantifying the influence of multiple meteorological factors with consideration for their lagged and regional impacts.

Methods

The researchers employed a multi-faceted methodological framework:

  • Multivariate Time Series Analysis: This was used to initially understand the temporal trends of Vibrio parahaemolyticus detection rates and their correlation with meteorological factors.
  • Principal Component Analysis (PCA): To address potential multicollinearity among the seven meteorological variables (temperature, humidity, precipitation, wind speed, sunshine duration, daily min/max temperature), PCA was used to reduce the dimensionality into three principal components representing temperature-related, water-related, and wind-related factors. This simplifies the analysis while retaining most of the variance.
  • Spatio-Temporal Scanning Statistics: This technique was crucial for identifying significant geographical and temporal clusters of Vibrio parahaemolyticus outbreaks. The study area was then stratified into these identified clusters and a non-clustered region.
  • Vector Autoregressive (VAR) Model: For each of the identified regions (three clusters and one non-clustered area), a VAR model was constructed. This model allows for the analysis of the dynamic relationships between multiple time series variables, specifically examining how the detection rate of V. parahaemolyticus responds to shocks in the principal meteorological components over time. This approach accounts for the interdependencies between variables and their lagged effects.
  • Granger Causality Test: Used to determine if past values of meteorological factors could predict future values of the disease detection rate.
  • Impulse Response Analysis: This was employed to visualize and quantify the dynamic impact of a shock in one variable (e.g., a change in temperature) on another variable (e.g., the detection rate) over time.

Datasets

The study utilized the following datasets:

  • Bacterial Foodborne Disease Data: Weekly case reports of Vibrio parahaemolyticus from January 2014 to December 2018 in Zhejiang Province, China. This data was collected from 101 sentinel hospitals covering 89 county-level jurisdictions.
  • Meteorological Data: Daily meteorological data from the China Meteorological Data Network for the same period and region. This included county precipitation, average wind speed, average temperature, average relative humidity, daily minimum temperature, daily maximum temperature, and sunlight hours. This data was interpolated using Inverse Distance Weighting (IDW) to county centroids.

No new benchmarks were introduced; the study focused on analyzing existing data with advanced statistical methods.

Key Insights and Ideas

  • Lagged Effects are Crucial: The most striking finding is that meteorological factors don't instantly impact disease rates. Temperature, humidity, and precipitation have significant lagging effects, meaning the influence of a weather event might not be seen for weeks. This is a critical insight for public health planning.
  • Regional Differences Matter: The study effectively demonstrates that the impact of meteorological factors on V. parahaemolyticus detection rates varies significantly between different spatio-temporal clusters. This challenges a one-size-fits-all approach to disease prevention.
  • Temperature is a Primary Driver: Temperature-related factors consistently showed a strong positive promoting effect on the detection rate across all regions, with a notable and sustained impact in the clustered areas.
  • Proactive Intervention is Key: The identified lag periods (e.g., 3 weeks for temperature, 8 weeks for humidity/precipitation) suggest that public health departments need to initiate prevention and response programs well in advance of anticipated climate conditions.

Key Results

  • Spatio-Temporal Clustering: Three significant spatio-temporal aggregation areas for vibriosis were identified in Zhejiang Province.
  • Seasonal Peak: Vibriosis incidence peaked in the summer months (June-August).
  • Geographical Hotspots: The detection rate was higher in eastern coastal areas and northwestern Zhejiang compared to the southwest.
  • Lagged Meteorological Influences:
    • Temperature: 3-week lag
    • Relative Humidity: 8-week lag
    • Precipitation: 8-week lag
    • Sunlight Hours: 2-week lag
    • Wind Speed: 11-week lag (weak negative correlation)
  • Regional VAR Model Performance: The VAR models for the four regions showed good performance, with adjusted R-squared values ranging from 0.573 to 0.727, indicating that the models explained a substantial portion of the variance in disease detection rates.
  • Impulse Response Findings: Temperature shocks led to a positive and sustained increase in detection rates, particularly in clustered areas. Moisture-related factors showed a more delayed and less pronounced impact.

Limitations and Future Work

The authors acknowledge several limitations:

  • Administrative Region Division: The study area was divided based on administrative regions and spatio-temporal scanning results, which might overlook differences in socio-economic conditions (e.g., per capita GDP disparities within non-clustered areas). Future work could explore more nuanced regional divisions.
  • Limited Factors Considered: The modeling primarily focused on meteorological factors. Future research should incorporate other influential indicators such as food consumption patterns, food exposure information, age structure, and financial/health expenditure to provide a more comprehensive understanding of foodborne disease drivers.

The proposed framework, however, offers a valuable approach for understanding spatio-temporal heterogeneity in disease dynamics and can be extended to other climate-sensitive public health issues.

Summary updated

Abstract

Foodborne diseases are a critical public health problem worldwide and significantly impact human health, economic losses, and social dynamics. Understanding the dynamic relationship between the detection rate of bacterial foodborne diseases and a variety of meteorological factors is crucial for predicting outbreaks of bacterial foodborne diseases. This study analyzed the spatio-temporal patterns of vibriosis in Zhejiang Province from 2014 to 2018 at regional and weekly scales, investigating the dynamic effects of various meteorological factors. Vibriosis had a significant temporal and spatial pattern of aggregation, and a high incidence period occurred in the summer seasons from June to August. The detection rate of <i>Vibrio parahaemolyticus</i> in foodborne diseases was relatively high in the eastern coastal areas and northwestern Zhejiang Plain. Meteorological factors had lagging effects on the detection rate of <i>V. parahaemolyticus</i> (3 weeks for temperature, 8 weeks for relative humidity, 8 weeks for precipitation, and 2 weeks for sunlight hours), and the lag period varied in different spatial agglomeration regions. Therefore, disease control departments should launch vibriosis prevention and response programs that are two to eight weeks in advance of the current climate characteristics at different spatio-temporal clustering regions.
BibTeX
@article{qi2023comprehensive,
  author = {Xiaojuan Qi and Jingxian Guo and Shenjun Yao and Ting Liu and Hao Hou and Huan Ren},
  title = {{Comprehensive Dynamic Influence of Multiple Meteorological Factors on the Detection Rate of Bacterial Foodborne Diseases under Spatio-Temporal Heterogeneity}},
  journal = {International Journal of Environmental Research and Public Health},
  year = {2023},
  doi = {10.3390/ijerph20054321},
}
Figure