Table 6.

Univariable and multivariable analysis of association between cardiovascular magnetic resonance imaging T1 and T2 maps and major adverse cardiovascular events.

	Univariable model		Multivariable model 1*		Multivariable model 2†
	HR (95% CI)	p Value	HR (95% CI)	p Value	HR (95% CI)	p Value
Native T1 value, per 1 unit increase in z score	1.31 (1.14, 1.50)	<0.001	1.54 (1.26, 1.87)	<0.001	1.44 (1.12, 1.84)	0.004 ‡
Native T1 value (1.5T Siemens), per 10ms increase	1.10 (1.03,1.17)	0.004	1.14 (1.05, 1.23)	0.001	1.11 (1.02, 1.21)	0.017 ¶
Abnormal T1 values§	-	-	-	-	-	-
T2 value, per 1 unit increase in z score	1.36 (1.12, 1.65)	0.002	1.32 (1.07, 1.61)	0.008	1.22 (0.98, 1.52)	0.077
T2 value (1.5T Siemens only), per 1ms increase	1.10 (1.00, 1.20)	0.042	1.07 (0.98, 1.18)	0.143	1.05 (0.96, 1.15)	0.312
Abnormal T2 values	2.86 (1.17, 6.99)	0.022	2.48 (1.00, 6.13)	0.049	2.19 (0.81, 5.91)	0.122

CMR = cardiovascular magnetic resonance

^*Multivariable Model 1: Cox proportional hazard model adjusting for age, sex.

^†Multivariable Model 2: Cox proportional hazard model adjusting for age, sex, number of cardiovascular risk factors, and LVEF by CMR during the index hospitalization.

^‡In sensitivity analysis when presence of LGE was added to multivariable model 2, the HR per 1 unit increase in z score remained significantly associated with MACE: HR 1.45 (95% CI: 1.14, 1,84), p=0.003.

^¶Similarly when LGE was added to multivariable model 2, native T1 values on the 1.5T magnet remained significanlty associated with MACE: HR per 10ms increase being 1.12 (95% CI: 1.02, 1.22), p=0.013. When assessing the association of LVEF by CMR with outcomes, LVEF was removed from the model.

^§Since all MACE occurred in patients abnormal with T1 value s (i.e. >mean+2SD) a HR could not be calculated.