Spatiotemporal Distribution of Hand, Foot, and Mouth Disease in Guangdong Province, China and Potential Predictors, 2009–2012

Abstract

Background: Hand, foot, and mouth disease (HFMD) is a common infectious disease among children. Guangdong Province is one of the most severely affected provinces in south China. This study aims to identify the spatiotemporal distribution characteristics and potential predictors of HFMD in Guangdong Province and provide a theoretical basis for the disease control and prevention. Methods: Case-based HFMD surveillance data from 2009 to 2012 was obtained from the China Center for Disease Control and Prevention (China CDC). The Bayesian spatiotemporal model was used to evaluate the spatiotemporal variations of HFMD and identify the potential association with meteorological and socioeconomic factors. Results: Spatially, areas with higher relative risk (RR) of HFMD tended to be clustered around the Pearl River Delta region (the mid-east of the province). Temporally, we observed that the risk of HFMD peaked from April to July and October to December each year and detected an upward trend between 2009 and 2012. There was positive nonlinear enhancement between spatial and temporal effects, and the distribution of relative risk in space was not fixed, which had an irregular fluctuating trend in each month. The risk of HFMD was significantly associated with monthly average relative humidity (RR: 1.015, 95% CI: 1.006–1.024), monthly average temperature (RR: 1.045, 95% CI: 1.021–1.069), and monthly average rainfall (RR: 1.004, 95% CI: 1.001–1.008), but not significantly associated with average GDP. Conclusions: The risk of HFMD in Guangdong showed significant spatiotemporal heterogeneity. There was spatiotemporal interaction in the relative risk of HFMD. Adding a spatiotemporal interaction term could well explain the change of spatial effect with time, thus increasing the goodness of fit of the model. Meteorological factors, such as monthly average relative humidity, monthly average temperature, and monthly average rainfall, might be the driving factors of HFMD.

Spatial and Temporal Characteristics of Hand-Foot-and-Mouth Disease and Its Response to Climate Factors in the Ili River Valley Region of China

“…We also used GAM to qualify the multiple climate factors affecting HFMD incidence. The associations were smaller in magnitude than those observed in other studies [ 26 , 27 ].…”

“…There were previous studies conducted in exploring spatiotemporal changes of HFMD and the effects of climate factors on this disease [ 27 , 42 ]. Regression models were commonly used to evaluate the relationships between possible influencing factors and the number of HFMD epidemic cases [ 43 , 44 ].…”

Spatial and Temporal Characteristics of Hand-Foot-and-Mouth Disease and Their Influencing Factors in Urumqi, China

“…This view was also supported by a study that found that GDP was significantly positively associated with the Shandong HFMD incidence [ 9 ]. A study conducted in Guangdong found that counties with higher relative risks were mainly gathered in the Pearl River Delta region, which is located in the economic center of Guangdong [ 10 ]. Another study concluded that urban areas had a higher risk of HFMD than poorer areas, which is similar to our finding that GDP is positively correlated with HFMD incidence [ 41 ].…”

“…Existing studies have widely accepted that socioeconomic factors and meteorological factors are related to the morbidity number of HFMD. The gross domestic product (GDP) [ 9 , 10 ] population density [ 11 , 12 ], road density [ 13 ] and land cover types [ 14 ] have also been demonstrated to affect the morbidity number of HFMD. Different HFMD incidence patterns have different climate conditions [ 15 ].…”

Using a Bayesian spatiotemporal model to identify the influencing factors and high-risk areas of hand, foot and mouth disease (HFMD) in Shenzhen

“…In the past, most studies have used time-series models (SIR, ARIMA) to estimate the impacts of variables [28][29][30]. In recent years, since the discovery that logistic regression cannot explain spatial heterogeneity and spatial correlation well, it is recommended that the Bayesian spatiotemporal model be used to identify spatiotemporal variations and the effects of potential predictors [31][32][33]. The method uses Markov chain Monte Carlo simulation to estimate the posterior distribution of random variables and to identify the relative risks in different districts.…”