Relationship between body composition and left ventricular geometry using three dimensional cardiovascular magnetic resonance

Table 2 Summary of the multiple linear regression models, split by gender

	Men			Women
	B	β	P	B	β	p	p for interaction
LV mass (g)
Lean Mass	1.91	0.53	<.0001	1.65	0.51	<.0001	.005
Fat Mass	−0.03	−0.01	.82	0.26	0.11	<.0001	.02
LV EDV (ml)
Lean Mass	1.80	0.49	<.0001	1.56	0.43	<.0001	.11
Fat Mass	−0.29	−0.07	.02	0.34	0.13	<.0001	<.0001
Concentricity (LV mass/volume)
Lean Mass	0.002	0.11	.03	0.004	0.19	<.0001	.65
Fat Mass	0.002	0.09	.03	−6×10⁻⁶	−0.0004	.99	.03
Stroke Volume (ml)
Lean Mass	1.01	0.44	<.0001	1.07	0.46	<.0001	.86
Fat Mass	−0.1	−0.04	.23	0.28	0.17	<.0001	<.0001
Heart Rate (bpm)
Lean Mass	−0.26	−0.20	.0003	−0.15	−0.10	.04	.48
Fat Mass	0.26	0.17	<.0001	0.09	0.08	.04	.04
Cardiac Output (L/min)
Lean Mass	0.03	0.18	.0002	0.06	0.29	<.0001	.19
Fat Mass	0.02	0.10	.006	0.03	0.20	<.0001	.09

Models are adjusted for age, race, systolic BP and height. Lean mass and fat mass are in kg
B gives the estimate of the beta-values in the regression equations, such that for each 1 kg increase in fat mass or lean mass the given predictor variable (e.g. LV mass) changes by B amount (if other variables in the model are held constant)
β gives standardised beta-values, such that for each 1 standard-deviation increase in lean mass or fat mass, the given predictor variable changes by β standard-deviations (if other variables in the model are held constant)
R ² for LV mass models: men = 0.42, women = 0.43. R ² for LV EDV models: men = 0.47, women = 0.45. R ² for concentricity models: men = 0.10, women = 0.10. R ² for stroke volume models: men = 0.42, women = 0.45. R ² for heart rate models: men = 0.06, women = 0.02. R ² for cardiac output models: men = 0.24, women = 0.29. BP indicates blood pressure, LV left ventricle and EDV end diastolic volume

ISSN: 1532-429X