Path diagram for bivariate latent growth curve modeling to estimate the effects of SBP and LDL on IMT within the same model. The three repeated measurements of systolic blood pressure (SBP1, SBP2, and SBP3), the three repeated measurements of LDL cholesterol (LDL1, LDL2, and LDL3), and carotid IMT (IMT) are in squares. F1 and F2 are the estimated baseline SBP and overall change in SBP, respectively; and F3 and F4 the estimated baseline LDL and changes in LDL, respectively. To simply the presentation, we did not show the residual error terms and the correlations among the later variables in the graph. The estimated coefficients and standard errors were shown and the significance level <0.05 was presented as ‘*’. Footnote: the estimated means±standard errors for F1 (SBP intercept) were 47.7±3.2, F2 (SBP change) were 17.2±3.3, F3 (LDL intercept) were 22.1±7.6, F4 (LDL change) were 23.6±5.4, all significance levels <0.05.
Chien et al. BMC Cardiovascular Disorders 2012 12:66 doi:10.1186/1471-2261-12-66