Biliary pancreatic diversion and laparoscopic adjustable gastric banding in morbid obesity: their long-term effects on metabolic syndrome and on cardiovascular parameters

Table 3 Linear and multiple regression

	Linear Regression	Multiple Regression
A
Δ Cornell Voltage Product		Model r = .635,
Δ BMI	.204, 0.0406
Cornell voltage product initial	.516, 0.0001	F = 45.471, 0.0001
Δ Systolic BP	.282, 0.0001	F = 9.616, 0.01
Δ Diastolic BP	.235, 0.0001
B
Δ RPP		Model r = .989,
Δ BMI	.297, 0.0025
RPP initial	.786, 0.0001	F = 158.97, 0.0001
Systolic BP	.491, 0.0001	F = 93.45, 0.0001
Diastolic BP	.387, 0.0001
HR	.616, 0.0001	F = 110.42, 0.0001
Δ HR	.754, 0.0001	F = 1721.69, 0.0001
Δ Systolic BP	.685, 0.0001	F = 1060.15, 0.0001
Δ Diastolic BP	.489, 0.0001

A. On the left: correlations (linear regression) between changes of Cornell voltage product (dependent variable) and changes of clinical variables (independent variables): r and p are indicated. On the right, multiple regression model, partial F and p of independent variables statistically significant. B. On the left: correlations (linear regression) between changes of RPP (dependent variable) and changes of clinical variables (independent variables): r and p are indicated. On the right, multiple regression model, partial F and p of independent variables statistically significant.

ISSN: 1475-2840