State Key Laboratory of Infectious Disease Prevention and Control, Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases, WHO Collaborating Centre for Vector Surveillance and Management, National Institute for Communicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention, Beijing, China.
Provincial Hospital Affiliated to Shandong First Medical University, Jinan, China.
Front Public Health. 2021 Jan 18;8:603872. doi: 10.3389/fpubh.2020.603872. eCollection 2020.
Determination of the key factors affecting dengue occurrence is of significant importance for the successful response to its outbreak. Yunnan and Guangdong Provinces in China are hotspots of dengue outbreak during recent years. However, few studies focused on the drive of multi-dimensional factors on dengue occurrence failing to consider the possible multicollinearity of the studied factors, which may bias the results. In this study, multiple linear regression analysis was utilized to explore the effect of multicollinearity among dengue occurrences and related natural and social factors. A principal component regression (PCR) analysis was utilized to determine the key dengue-driven factors in Guangzhou city of Guangdong Province and Xishuangbanna prefecture of Yunnan Province, respectively. The effect of multicollinearity existed in both Guangzhou city and Xishuangbanna prefecture, respectively. PCR model revealed that the top three contributing factors to dengue occurrence in Guangzhou were Breteau Index (BI) (positive correlation), the number of imported dengue cases lagged by 1 month (positive correlation), and monthly average of maximum temperature lagged by 1 month (negative correlation). In contrast, the top three factors contributing to dengue occurrence in Xishuangbanna included monthly average of minimum temperature lagged by 1 month (positive correlation), monthly average of maximum temperature (positive correlation), monthly average of relative humidity (positive correlation), respectively. Meteorological factors presented stronger impacts on dengue occurrence in Xishuangbanna, Yunnan, while BI and the number of imported cases lagged by 1 month played important roles on dengue transmission in Guangzhou, Guangdong. Our findings could help to facilitate the formulation of tailored dengue response mechanism in representative areas of China in the future.
确定影响登革热发生的关键因素对于成功应对登革热疫情至关重要。中国的云南省和广东省是近年来登革热爆发的热点地区。然而，很少有研究关注多维因素对登革热发生的驱动作用，且未能考虑到所研究因素之间可能存在的多重共线性，这可能会导致结果出现偏差。在本研究中，我们利用多元线性回归分析探讨了登革热发生及其相关自然和社会因素之间的多重共线性影响。我们还利用主成分回归（PCR）分析分别确定了广东省广州市和云南省西双版纳傣族自治州的关键登革热驱动因素。结果表明，广州市和西双版纳傣族自治州均存在多重共线性的影响。PCR 模型显示，广州市登革热发生的前三个主要影响因素为布雷图指数（BI）（正相关）、滞后 1 个月的输入性登革热病例数（正相关）和滞后 1 个月的月平均最高温度（负相关）。相比之下，导致西双版纳傣族自治州登革热发生的前三个因素分别为滞后 1 个月的月平均最低温度（正相关）、月平均最高温度（正相关）和月平均相对湿度（正相关）。气象因素对云南省西双版纳傣族自治州的登革热发生影响更强，而 BI 和滞后 1 个月的输入性病例数在广东省广州市的登革热传播中发挥着重要作用。本研究结果有助于未来在我国有代表性的地区制定有针对性的登革热应对机制。
