基于logistic回归模型的儿童过敏性紫癜影响因素研究

Influencing factor study of Henoch-Schonlein purpura in children based on the logistic regression model

目的：分析儿童过敏性紫癜的影响因素，为疾病的预防提供线索和依据。方法通过调查问卷收集414名过敏性紫癜及对照组儿童的相关信息；分别采用t检验、χ2检验和秩和检验比较病例组和对照组的差异。利用logistic回归模型分析儿童过敏性紫癜的影响因素。结果单因素分析结果显示，病例组和对照组年龄、母亲和父亲受教育程度差异有统计学意义；饮食规律性、食用生冷、油炸和辛辣食物、肉类、饮料、奶及奶制品、水果和蔬菜差异有统计学意义；出生方式、母乳喂养、混合喂养以及母亲孕前接触有害物质差异有统计学意义。调整年龄、性别和母亲受教育程度后多因素分析结果显示，儿童过敏性紫癜的保护因素有：生活在农村地区、睡眠时间、水果食用频率、甜食食用频率、肉类食用频率、母亲受教育程度；儿童过敏性紫癜的危险因素有：混合喂养、口嚼食物喂养、怀孕期间患病、生冷食物食用频率、坚果食用频率、辛辣食物食用频率。结论保证儿童充足的睡眠、少吃生冷和辛辣食物以及充足的肉类和水果可以降低过敏性紫癜的发生风险；预防过敏性紫癜还可以提倡母乳喂养、避免口嚼食物喂养。

Objective The aim of this study is to assess the associated factors of Henoch-Schonlein purpura（ HSP） in Chinese children. This study has important public implications for developing strategies of preventing HSP. Methods In this hospital-based, case-control study, we recruited 353 HSP cases and 61 control participants during 2012 and 2015. We collected related information of 414 HSP in children and the control group through questionnaires. Student＇s t-test, Pearson Chi-square test and Wilcoxon test were used respectively to compare the difference between the case and control group; We applied logistic re-gression model to analyze the associated factors of HSP in children. Results Univariate analysis revealed that the differences in age, mother and father＇s level of education were significant between the case and control group. Univariate analysis also revealed that the differences in diet regularity, eating cold, fried and spicy food, meat, drinks, milk and dairy products, fruits and vege-tables between the two groups were statistically significant. Similarly, the differences in way of birth, breast-feeding, mixed feed-ing and mother contacting harmful substances before becoming pregnant between the two groups were statistically significant. After the adjustment of age, sex and mother＇s level of education, multivariate Logistic regression demonstrated that living in rural are-as, the enough sleep time, consumption frequencies of fruit, sweet and meat and mother＇s level of education were protective fac-tors for HSP in children. Multivariate Logistic regression also demonstrated that mixed feeding, feeding after chewing, illness during pregnancy, and consumption frequencies of cold, spicy food and nuts were the risky factors for HSP. Conclusion Suffi-cient sleep time, less cold and spicy food and plenty of meat and fruit can reduce the risk of HSP in children. Advocating breast-feeding, avoid feeding after chewing can also prevent children from HSP.