Table 2.

Multivariate analyses of LVMI and LVEFa

Independent VariablesβP
aLVMI, left ventricular mass index; LVEF, left ventricular ejection fraction; ANP, atrial natriuretic peptide; ln, logarithm (natural); BNP, brain natriuretic peptide. Because of the strong ANP-BNP collinearity (see Results) the predictive power of the two peptides for LVMI and LVEF was tested in separate multiple regression models. Data are standardized regression coefficient (β) and P values. Factors that did not contribute significantly to the models may be deducted from the description of the multivariable strategy adopted in this study (see Statistical Analyses section).
bMultiple R = 0.64; P = 0.0001.
cMultiple R = 0.63; P = 0.0001.
dMultiple R = 0.49; P = 0.0001.
eMultiple R = 0.52; P = 0.0001.
LVMI
ANP-based modelb
ln ANP0.330.0001
ln albumin-0.210.0001
systolic pressure0.160.006
age0.150.009
antihypertensive therapy0.120.03
diabetes mellitus0.110.04
ln iPTH0.100.05
BNP-based modelc
ln BNP0.370.0001
ln albumin-0.190.001
systolic pressure0.160.002
age0.160.003
diabetes mellitus0.110.04
LVEF
ANP-based modeld
ln ANP-0.320.0001
gender0.180.002
ln iPTH-0.180.002
diabetes mellitus-0.160.005
antihypertensive therapy-0.130.02
BNP-based modele
ln BNP-0.400.0001
gender0.200.0001
diabetes mellitus-0.180.001
ln iPTH-0.180.001