Multivariate response model with multilevel and its application in the influencing factors of blood pressure
Yang, Y.; Fu, P.; Xie, J.; Zhang, W.; Zhang, M.; Wang, C.; Ping, Z.; Hu, D.
Wei Sheng Yan Jiu 38(5): 582-584
2009
ISSN/ISBN: 1000-8020 PMID: 19877521 Document Number: 631203
To explore the application of multivariate response model with multilevel in the influencing factors of blood pressure. Two response model with three-level was fitted under MLwin 2.02 software. The correlation coefficient between systolic blood pressure (SBP) and diastolic blood pressure (DBP) was 0.949 at region level, and 0.701 at individual level. SBP and DBP level increased with age, while the regression coefficient of age on SBP was significantly higher than on DBP, beta was 0.720 (SBP) and 0.118 (DBP) individually (chi2 = 4284.56, P < 0.001). The DBP and SBP level of male were higher than that of female, while the regression coefficient of gender on DBP was significantly higher than on SBP, beta was 2.208 (SBP) and 3.113 (DBP) individually (chi2 = 31.35, P < 0.001). Multivariate response model with multilevel can be used to analyze the hierarchy structure data, and it is also a good tool to analyze the influencing factors of blood pressure.