A: OLS regression coefficients for the regression of ln(PWV) onto sex, age, duration of DM, and HbA1c | |||||
Regression parameter | Parameter estimate | SE | Lower 95% CL | Upper 95% CL | |
β0 | 0.9896 | 0.3014 | 0.3749 | 1.6043 | |
β1 | −0.1435 | 0.0681 | −0.2824 | −0.0046 | |
β2 | 0.0190 | 0.0074 | 0.0039 | 0.0341 | |
β3 | −0.0051 | 0.0073 | −0.02 | 0.0098 | |
β4 | 0.0545 | 0.0299 | −0.0065 | 0.1155 | |
B: Type III ANOVA F-tests | |||||
Predictor variable | df | Partial sum of squares | Mean square error | F-statistic | P value |
Sex | 1 | 0.16 | 0.16 | 4.44 | 0.043 |
Age | 1 | 0.24 | 0.24 | 6.57 | 0.015 |
Duration DM | 1 | 0.02 | 0.02 | 0.49 | 0.489 |
HbA1c | 1 | 0.12 | 0.12 | 3.32 | 0.078 |
Total | 4 | 0.95 | 0.24 | 6.46 | 0.001 |
Error | 31 | 1.14 | 0.04 |
(A) presents OLS regression coefficients for the OLS regression model: E(ln(cfPWV)|X)=β0+β1(female)+β2(age)+β3(duration of DM)+β4(HbA1c). (B) presents type III ANOVA F-tests of the OLS analysis.
ANOVA, analysis of variance; cfPWV, carotid-femoral pulse wave velocity; DM, diabetes mellitus; HbA1c, hemoglobin A1c; OLS, ordinary least squares.