天津地区脑出血危险因素的多元回归分析——96对配比病例对照研究

对96对脑出血病例及对照进行配比分析.经条件Logistic回归分析,发现以下因素与脑出血有关:高血压史、脑出血家族史、同胞高血压史、喜食肥肉.通过将高血压患病病程、血压值、血糖做为连续变量引入条件Logistic模型,发现脑出血发病的OR值与三个连续变量呈剂量反应关系.同时发现血压不稳定者易发生脑出血.还有些因素如冠心病史虽单因素分析结果有意义,但经高血压病史调整后,冠心病对脑出血的作用失去统计学意义.说明冠心病与脑出血的关系是通过高血压起作用的.

THE MULTIPLE RECESSION ANALYSIS ON RISK FACTORS OF CEREBRAL HEMORRHAGE

In the article 96 cases with cerebral hemorrhage and controls were studied by using the matched case-control study. Some factors were found to be related to cerebral hemorrhage by conditional logistic regression analysis: hypertension (OR= 6.88), cerebral hemorrhage family history(ORfather=8. 00 ORmother = 9. 00 ORsiblings = 9. 00), hypertension history of siblings (OR = 4. 00) taste for salt(OR= -0. 65) ,the type of eating pork(OR = 2. 40). Significant relationship between some spiritual factors and cerebral hemorrhage was found. By leading of the course of hypertension, the levels of blood pressure and glucose in serum as continuous variales into the conditional logistic regression model we found that there were dose-response relationships between those and cerebral hemorrhage. The significant relationship was also observed between unstable blood pressure and cerebral hemorrhage. The association between coronary heart disease and cerebral hemorrhage in simple analysis(OR = 2. 57)became not hgnificant after adjusting for hypertension. It suggested that hypertension might have the effect on the relationship between coronary heart disease and cerebral hemorrhage.