摘要
目的查找研究猩红热发病率和气象因素之间关系的适用方法,探讨邯郸市猩红热的气象流行病学特征。方法收集1972-2010年邯郸市猩红热疫情资料、气象资料和人口资料,采用EpiData3.0进行"双重录入",用SPSS17.0统计分析软件建立数据库,对数据进行统计分析。结果①气象参数的共线性诊断结果显示,本组气象因子数据容差最小为0.014,方差膨胀因子最大达69.998。②Spearman相关分析结果显示,邯郸市1972-2010年猩红热月发病率与月平均风速、月日照时数、月小型蒸发量呈正相关,与月平均气温、月平均相对湿度、月总降雨量、月极端最低气温呈负相关(P<0.05或P<0.01)。③猩红热月发病率的曲线估计方程为Y=1.369-0.230ln(X)。④猩红热月发病率与月平均风速之间得到曲线拟合方程Y=-0.781+1.242X-0.585X2+0.097X3。⑤气象参数的KMO和Bartlett检验结果显示,本文中的气象参数非常适合做因子分析,通过做主成分多元线性回归分析得到方程Y=1.946+0.378Z2(P<0.01)。结论①邯郸市10个气象参数之间存在严重的多重共线性。②邯郸市猩红热月发病率的模型曲线为对数模型曲线。③猩红热月发病率与月平均风速之间呈三次方程曲线关系,月平均风速是影响猩红热月发病率的主要气象因素。④气象因素对猩红热发病的影响在总的影响因素中所占比例较小。
Objective To look for the scientific method which could apply to research the relationship between the incidence of scarlet fever and meteorological factors, and to discuss the meteorological and epidemiological characteristics of scarlet fever in Handan City. Methods The data on scarlet fever, meteorological parameters and population in Handan City in 1997- 2010 were collected and inputted doubly to EpiData 3.0. The database was established using SPSS17.0 statistical analysis software, and then the data were analyzed. Results Collinearity diagnosis of meteorological parameters showed that the minimal toler- ance was 0. 014 and the maximal variance inflation factor was 69. 998. The results of Spearman＇ s correlation analysis showed that the monthly incidence of scarlet fever had significantly p^sitive correlation with monthly average wind speed, monthly sunshine hours, and monthly small - scale evaporation capacity, but had significantly negative correlation with monthly average tempera- ture, monthly average relative humidity, monthly rainfall, and monthly extreme minimum temperature （ P 〈 0.05 or P 〈 0.01）. The curve estimating equation of scarlet fever incidence was ~＂ = 1. 369- 0. 230In（X）. The curve fitting equation of monthly incidence and monthly average wind speed of scarlet fever was ＂~＂ = - 0. 781 ＋ 1. 242X - 0. 585X2 ＋ 0. 097X3. The re- sults of KMO and Bartlett＇s tests of meteorological parameters showed that the parameters were very suitable for factor analysis. The equation was obtained by principal component multivariate linear regression analysis, ＂~ = 1. 946 ＋ 0. 37874 （P〈 0.01 ）. Conclusions The problem of multicollinearity in 10 meteorological parameters in Handan City is very serious. The model curve of the monthly incidence of scarlet fever in Handan City is a logarithmic model curve. The relatioilship between the incidence of scarlet fever and the monthly average wind speed is a cubic equation curve, and the monthly average wind speed is the main mete- oro
出处
《实用预防医学》
CAS
2014年第2期194-196,共3页
Practical Preventive Medicine
关键词
猩红热
气象因素
气象流行病学
曲线估计
曲线拟合
因子分析
Scarlet fever
Meteorological factor
Meteorological epidemiology
Curve estimation
Curve fitting
Factor analysis
