Table 6

Mediation by HS-SIC on the association of risk factors with incident HF

	Mediation by high intensity signal intensity coefficient
	Direct effect			Mediated effect** (ie, indirect effect)			Total effect
	Estimate	SE	P value	Estimate	SE	P value	Estimate	SE	P value
Age	0.0359	0.0048	<0.001	0.0014	0.0007	0.053	0.0374	0.0048	<0.001
Sex	−0.0206	0.0086	0.010	−0.0054	0.0020	0.010	−0.0260	0.0082	0.001
Body mass index	0.0106	0.0043	0.014	0.0038	0.0016	0.019	0.0145	0.0044	0.001
Total/HDL cholesterol	0.0028	0.0045	0.54	0.0024	0.0009	0.007	0.0052	0.0043	0.24
Systolic blood pressure	0.0128	0.0042	0.002	0.0013	0.0005	0.012	0.0141	0.0041	0.001
Antihypertensive treatment	0.0407	0.0083	<0.001	0.0036	0.0015	0.017	0.0442	0.0083	<0.001
Diabetes	0.0538	0.0172	0.002	0.0060	0.0025	0.019	0.0598	0.0172	0.001
Smoking	−0.0244	0.0093	0.007	−0.0012	0.0010	0.26	−0.0256	0.0093	0.005

*The mediated effect can be considered the proportion of the total magnitude of association (ie, total effect) attributable to the mediator, in this case HS-SIC. For example, results shown in the table above suggest that HS-SIC is a significant mediator of the association between diabetes and incident HF, whereby 10% (indirect effect/total effect=0.006/0.060=0.10) of the total magnitude of risk observed for diabetes in relation to HF (total effect=0.060) is attributable to an increase in HS-SIC (indirect effect=0.006).
HDL, high density lipoprotein; HF, heart failure; HS-SIC, high spectrum signal intensity coefficient.