Zhang Han, Si Yali, Wang Xiaofeng, Gong Peng
Ministry of Education Key Laboratory for Earth System Modeling, Department of Earth System Science, Tsinghua University, Beijing 100084, China.
Joint Center for Global Change Studies, Beijing 100875, China.
Int J Environ Res Public Health. 2017 Jul 14;14(7):782. doi: 10.3390/ijerph14070782.
Bacillary dysentery has long been a considerable health problem in southwest China, however, the quantitative relationship between anthropogenic and physical environmental factors and the disease is not fully understand. It is also not clear where exactly the bacillary dysentery risk is potentially high. Based on the result of hotspot analysis, we generated training samples to build a spatial distribution model. Univariate analyses, autocorrelation and multi-collinearity examinations and stepwise selection were then applied to screen the potential causative factors. Multiple logistic regressions were finally applied to quantify the effects of key factors. A bootstrapping strategy was adopted while fitting models. The model was evaluated by area under the receiver operating characteristic curve (AUC), Kappa and independent validation samples. Hotspot counties were mainly mountainous lands in southwest China. Higher risk of bacillary dysentery was found associated with underdeveloped socio-economy, proximity to farmland or water bodies, higher environmental temperature, medium relative humidity and the distribution of the Tibeto-Burman ethnicity. A predictive risk map with high accuracy (88.19%) was generated. The high-risk areas are mainly located in the mountainous lands where the Tibeto-Burman people live, especially in the basins, river valleys or other flat places in the mountains with relatively lower elevation and a warmer climate. In the high-risk areas predicted by this study, improving the economic development, investment in health care and the construction of infrastructures for safe water supply, waste treatment and sewage disposal, and improving health related education could reduce the disease risk.
长期以来,细菌性痢疾一直是中国西南部一个相当严重的健康问题,然而,人为因素和自然环境因素与该疾病之间的定量关系尚未完全明确。细菌性痢疾潜在高风险的确切位置也不清楚。基于热点分析结果,我们生成了训练样本以构建空间分布模型。然后应用单变量分析、自相关和多重共线性检验以及逐步选择来筛选潜在的致病因素。最后应用多重逻辑回归来量化关键因素的影响。在拟合模型时采用了自助法策略。通过受试者工作特征曲线下面积(AUC)、kappa值和独立验证样本对模型进行评估。热点县主要位于中国西南部的山区。发现细菌性痢疾的较高风险与社会经济欠发达、靠近农田或水体、环境温度较高、相对湿度中等以及藏缅族的分布有关。生成了一幅准确率较高(88.19%)的预测风险图。高风险地区主要位于藏缅族居住的山区,特别是海拔相对较低、气候较温暖的山区盆地、河谷或其他平坦地区。在本研究预测的高风险地区,改善经济发展、加大医疗保健投入、建设安全供水、废物处理和污水处理基础设施以及加强健康相关教育可以降低疾病风险。