【作者】 付青；

【导师】 汤林华；

【作者基本信息】 中国疾病预防控制中心 ， 流行病与卫生统计学， 2013， 博士

【摘要】 新疆维吾尔自治区是目前我国内脏利什曼病的主要流行区之一,其中克孜勒河流域的喀什、疏附、疏勒、伽师4县(市)为重流行区,有人源型和自然疫源型内脏利什曼病的流行。内脏利什曼病属于被忽视的传染病之一。既往工作主要围绕着媒介白蛉的分类鉴定、地域分布和生态学等方面开展,鲜见与该病流行规律、传播特征有关的研究,在白蛉分子生物学方面的研究也有限。近年来,随着空间统计学和分子生物学方法、技术的广泛应用,为从不同角度研究我国内脏利什曼病流行特征及白蛉群体遗传结构提供了新手段并具有重要现实意义。本研究收集了克孜勒河流域4县(市)逐乡(镇)、逐月的发病数据、人口数据、地理空间数据、环境因素数据以及部分乡(镇)的媒介白蛉数据。通过运用流行病学、空间统计学、分子生物学的方法对研究地区内脏利什曼病发病的时间-空间聚集性、发病热点的时空演变、疾病流行的环境因素、媒介白蛉的分子群体遗传结构进行了研究,获得以下主要结果：一、内脏利什曼病发病的时间-空间聚集性运用空间统计分析扫描统计量法对各乡(镇)逐年的空间聚集性进行扫描,获得16次内脏利什曼病发病的空间聚集(P均<0.01),可归为2个一级聚集区和1个二级聚集区。对连续13年的数据进行时空聚集性扫描,获得2个内脏利什曼病发病的高危聚集区及其对应的高风险时段,2个聚集区内的内脏利什曼病发病风险分别是其他相同人口半径、相同时间间隔的扫描窗口发病风险的27.34倍和29.09倍(P均<0.01)二、内脏利什曼病发病热点的时空演变运用空间统计方法对研究地区内脏利什曼病的发病进行地理区域相关性分析,Global Moran’s I=0.18, Z Score=3.89(P<0.01)、General G=0.59, Z Score=4.11(P<0.01),结果说明内脏利什曼病的发病具有空间自相关性,且为高值聚集。运用地统计分析模块建立发病预测图并进行热点分析,累计探测到80个发病热点,内脏利什曼病在局部区域呈现中度甚至重度的聚集性,发病热点随时间的推移变化明显,地理位置表现为热点从中部-东部-中部的转移趋势,该病类型表现出从人源型-自然疫源型-人源型的转移趋势。三、内脏利什曼病流行的环境因素研究运用多元线性回归筛选与年发病率(1/万)相关的环境变量,建立不同类型内脏利什曼病流行的疾病-环境模型,定量分析了环境相关因素对该病流行的影响。拟合的人源型和自然疫源型内脏利什曼病模型分别见①和②：①√INCIDENCE=-8.64+0.35QTM6Z8+9.46QEM6Z8-0.06QRM1Z3经过筛选,纳入模型的环境变量有3个,分别为：前一年6-8月平均地面温度、前一年6-8月平均EVI、前一年1-3月平均地面降水。模型的决定系数R2=0.61,F=12.67,P<0.01。②√INCIDENCE=-16.05+0.29T6+0.82QTMYEAR-0.21QRM1Z3经过筛选,纳入模型的环境变量有3个,分别为：当年6月平均地面温度、前一年全年平均地面温度、前一年1-3月平均地面降水。模型的决定素数R2=0.34,F=4.13,P<0.05。四、媒介白蛉的分子群体遗传结构研究应用mtDNA-Cytb基因研究了吴氏白蛉和长管白蛉的群体遗传结构。结果显示FST的范围最小值在长管白蛉群体间,最大值在吴氏白蛉与长管白蛉群体间,吴氏白蛉群体间FST均为负值,提示群体间的遗传差异非常小。吴氏白蛉(R2=0.34,P<0.05)与长管白蛉(R2=0.09,P<0.05)基因流与地理距离均呈负相关,群体遗传结构格局符合距离隔离模型。结论：克孜勒河流域内脏利什曼病流行存在2个一级聚集区和1个二级聚集区。2个一级聚集区分别是以绿洲生态型为特征的人源型聚集区和以荒漠生态型为特征的自然疫源型聚集区。构建了该地区内脏利什曼病的流行的疾病-环境模型,该病的传播与特定时间的降水和温度呈相关关系。媒介白蛉的分子群体遗传结构研究揭示了该地区不同蛉种的种群遗传特征,为内脏利什曼病的疫情评估、监测预警提供了新的思路。更多还原

【Abstract】 Xinjiang Uygur Autonomous Region is among the Visceral Leishmaniasis (VL) endemic areas in China, where the disease is most serious in4counties along the Kezile River, i.e, Kashi, Shufu, Shule and Jiashi. As one of the neglected tropical diseases (NTD), VL represents2types, anthroponotic VL (AVL) and desert sub-type of zoonotic VL (DST-ZVL) in the region. A lot of efforts were made on classification of the vector, namely sandfly, geographical distribution and ecological researches on sandfly. Nevertheless, large gaps still remain in the nature of epidemiology, transmission of VL, as well as molecular biological research on sandfly. In recent years, the advance of spatial statistics and molecular biology have allowed for deepening our understanding on the epidemiology and the genetics of sandfly population from a newly different view.In the study, monthly data were collected from the4counties, including epidemiological, demographic, geographic, and environmental data. Sandflies were collected from some of those townships. Spatio-temporal clustering, evolution on hot spots of VL incidence, and environmental impact factors of VL, in addition to sandfly population genetic structure were studied through epidemiology, spacial statistics and molecular biology. Main outputs are presented as follows.1. Spatio-temporal clustering analysis of VLBased on scan statistic, a total of16VL spatial clusters by township by year were obtained (P<0.01), which could be classified as2most likely significant clusters and1significant secondary cluster. Spatio-temporal clustering analysis in13consecutive years resulted in2high risk clustering zones with the corresponding high risk time frames. It showed VL risk in the2clusters was27.45and29.09 times higher than that other zones with the same population radius and time interval (P<0.01).2. Evolution of hot spot of VL incidenceCorrelativity of geographic region showed the score of Global Moran’s/and Z value was0.18and3.89(P<0.01), the Local General G and Z value was0.59and4.11(P<0.01), respectively. It indicated that VL distribution was of high spatial autocorrelation with high-high clustering. By using geostatistical analysis, prediction map was illuminated and hot spot analysis was conducted. In general,80hot spots were identified, which showed medium or highly clustering of VL incidence in some certain areas. In addition, the distribution of hot spots showed a moving tendency of middle-east-middle in geography, while the disease distribution showed a tendency of AVL-DSTZVL-AVL with obvious variations of hot spots over time.3. Environmental impact factors of VLBased on correlation analysis and the multivariable linear regression model, environmental variables relevant to the incidence of VL were screened to set up the disease-environment model for different VL types. The impacts of certain environmental factors on VL incidence were then quantified. Two models for AVL (see①) and DST-ZVL (see②) are described as follows.①INCIDENCE=-8.64+0.35QTM6Z8+9.46QEM6Z8-0.06QRM1Z3R2=0.61,F=12.67, P<0.01.②INCIDENCE=-16.05+0.29T6+0.82QTMYEAR-0.21QRM1Z3R2=0.34, F=4.13, P<0.05.4. Molecular population genetic structure of sandflyMolecular population genetic structures of Phlebotomus wui and Phlebotomus Longiductus were studied by using mtDNA-Cytb. Results showed the lowest FST was in the P. Longiductus population while the highest was among both two species. The negative FST of P. wui indicated minor genetic difference among this population. The level of gene flow of P. wui (R2=0.34, P<0.05) and P. Longiductus (R2=0.09, P<0.05) showed negatively relative with geographic distance, and the pattern of population genetic structure was in accordance with isolation-by-distance model.更多还原