. Author manuscript; available in PMC: 2009 Jun 24.

Published in final edited form as: J Am Coll Cardiol. 2008 Jun 24;51(25):2414–2421. doi: 10.1016/j.jacc.2008.03.018

Table 3.

Cox proportional hazards analysis for the time to occurrence of an index composite cardiac event.

Variable	UNIVARIATE ANALYSIS		MULTIVARIATE ANALYSIS MODEL 1^*		MULTIVARIATE ANALYSIS MODEL 2^†
	UNADJUSTED HR VALUE (95% CI)	P-	ADJUSTED HR VALUE (95% CI)	P-	ADJUSTED HR VALUE (95%CI)	P-
Age (5 year increments)	1.1 (0.9–1.5)	0.28	---	---	---	---
Gender	1.8 (0.5–6.2)	0.39	---	---	---	---
NYHA functional class	2.0 (0.8–4.5)	0.12	2.7 (1.0–7.2)	0.04	2.6 (1.0–7.0)	0.05
CMR LVEDV index (ml/m²) (per 10 ml/m² increments)	1.1 (1.0–1.2)	0.08	---	---	1.0 (0.9–1.1)	0.70
Presence of CMR LGE (vs absence)	7.1 (2.0–25.4)	0.002	8.6 (2.4–31.6)	0.001	8.2 (2.2–30.9)	0.002
Excluding patients with CAD
Age (5 year increments)	1.2 (0.9–1.6)	0.17	---	---	---	---
Gender	1.9 (0.5–6.8)	0.33	---	---	---	---
NYHA functional class	1.8 (0.8–4.4)	0.17	---	---	2.1 (0.8–5.5)	0.15
CMR LVEDV index (ml/m²) (per 10 ml/m² increments)	1.1 (1.0–1.2)	0.14	---	---	1.0 (0.9–1.1)	0.9
Presence of CMR LGE (vs absence)	8.9 (2.4–32.2)	0.001	8.9 (2.4–32.2)	0.001	9.2 (2.4–35.1)	0.001

For Model 1 (parsimonious model), forward stepwise regression (p<0.10 for entry, p>0.05 for removal) was used to arrive at a parsimonious model.

^†

For Model 2, NYHA functional class, CMR LVEDV index, and LGE fibrosis were all entered into the multivariate model.