Development of Predictive Models for Foodborne Disease Outbreak Due to Climate Change in Korea

Introduction:  Climate change may have direct and indirect impact on the food safety at various stages of the food chain in Korea. Especially, it is expected to cause an increase in the foodborne disease outbreaks (FBDOs). Therefore their corresponding studies have been actively conducted in many country. However accurate prediction of FBDOs requires considering the differences of pattern in FBDOs between countries and regions and developing a model by appropriately utilizing the major variables of climate change.

Purpose:  Therefore, in this study, to identify the impact of climate change on FBDOs in Korea, predictive models were being developed using climate variables in consideration of regional differences of Korea. Also, future trends of FBDOs due to climate change in Korea were predicted by the mathematical model combining climate change scenario.

Methods:  We analyzed datasets on FBDOs together with 18 climate variables in Korea during the period 2002-2011, and developed the Poisson GLM, ARMA, and Auto-Reg models. Notified FBDOs in 2011 were used to test the predictive ability of the models. Parameter estimation, goodness-of-fit and predictive ability (AIC, BIC, and MSE) of the models were compared. For the future trends of FBDOs, the climate change scenario produced in KMA (Korean Meterological Administration) based on the SERS A1B emission scenarios were used.

Results:  The results suggested that the Auto-Reg and ARMA models produced the highest predictive ability; however, Poisson GLM and Poisson AutoRegressive models produced significantly large MSEs, indicating relatively lower predictive abilities than the other models. Though the high relationship was shown between meteorological variables, average temperature was all positively correlated with FBDOs in three climatic regions classified by temperature and rainfall patterns in Korea. The impact of climate change will exacerbate incidences of foodborne disease in Korea, so until 2040 about 1.3 (outbreaks) and 1.5 (cases) fold increase with regional and seasonal variation.

Significance:  This is the first study to examine the association between climate variability and FBDOs using different predictive models considered regional variations in Korea. The predictive models play an essential tool for developing food safety programs and climate change adaptation in Korea.