|
The latest statistical methods for the purpose of this project is a thorough risk factors of CI incidence and its evaluation on a variety of methods to compare and evaluate. Methods A 1:1 matched case - control study, 150 cases in each collection of cases and controls, to investigate past medical history, family history, environmental factors, dietary, nutritional factors epidemiological data. Logistic regression model, path analysis, linear structural equation model of the factors affecting CI incidence study and evaluation of these methods, from a methodological point of view. Results (1) Logistic regression analysis reveal the activities of a small amount of physical exercise, eating fresh fruits, fresh seafood, fresh beans, fresh pickles, history of hypertension, economic pressure, triglycerides impact of factors affecting the CI incidence, which physical exercise (OR = 0.05; 95% CI ,0.020-0 .127; P <0.0001), consumption of fresh fruit (OR = 0.185; 95% CI, between 0.07 and 0 .47; P = 0.0004), fresh seafood (OR = 0.346; 95 % CI ,0.13-0 .89; P = 0.0279), food legumes (OR = 0.238; 95% CI ,0.116-0 .487; P <0.0001) to as CI protective factors, history of hypertension (OR = 4.3; 95% CI, 1.8-10.4; P = 0.0011), the revenue pressure (OR = 3.516; 95% CI, 1.4 - 8 .9; P = 0.0079), activity (OR = 3.85; 95% CI ,1.46-10 .12; P = 0.0063), Fresh pickles (OR = 3.437; 95% CI ,1.28-9 .26; P = 0.0146), triglycerides (OR = 3.787; 95% CI ,1.57-9 .12; P = 0.0030) as risk factors. (2) path analysis showed that: hypertension, fresh seafood, fresh fruits, fresh beans, triglycerides CI the dish Nankou Wyatt School stuffy bite d Lun Wen dagger degree this has a direct impact, economic pressure, salt, Sports exercise, less activity through each intermediate links work. (3) linear structural equation model (LISREL) analysis showed that, Fl factor CI total effect weight coefficient to .4051: FZ factor through F3 factor indirectly affect CI, effect the right to re-coefficient of a 0.039; F3 factor directly effect the right weight coefficient to 0.3000 ; F4 through F3 factor indirectly affect the disease, affecting the value of the effect 0.1290; F5 through F6 factor indirectly affect CI incidence the impact weight coefficient 0.0615; F6 factor directly affects Cl weight coefficient of 0.4100. Conclusion (1) (2) (3) Logsitic regression analysis combined with path analysis, the direct and indirect effects of various factors can be calculated, can only be used for Effects of significant variables related to the latent variable powerless; linear structural equation better way to solve the problem of latent variable analysis, enhance researchers' understanding of the deep-seated problems of causality. The LlsREL able than Logsitic model to analyze impact factors to better dig Cl incidence of influencing factors.
|