Table 3.
Cox Proportional Hazards Models for Incident Cardiovascular Events as a Function of RMD
| Clinical Endpoint | RMD Adjusted HR (95% CI) | |||
|---|---|---|---|---|
| Model 1* | Model 2† | Model 3‡ | Model 4§ | |
| Heart failure | 2.62 (1.56–4.39); p < 0.001 | 2.52 (1.49–4.26); p = 0.001 | 1.87 (1.06–3.30); p = 0.030 | 1.80 (1.02–3.20); p = 0.044 |
| Hard coronary events (MI, resuscitated cardiac arrest, and death from CAD) | 1.76 (1.10–2.81); p = 0.018 | 1.76 (1.09–2.85); p = 0.021 | 1.82 (1.10–2.99); p = 0.019 | 1.75 (1.06–2.89); p = 0.029 |
| All coronary events (hard coronary events, definite angina, probable angina followed by revascularization) | 1.56 (1.10–2.21); p = 0.012 | 1.56 (1.09–2.22); p = 0.014 | 1.55 (1.07–2.24); p = 0.020 | 1.52 (1.05–2.21); p = 0.026 |
| Hard cardiovascular events (hard coronary events, nonfatal and fatal stroke) | 1.70 (1.17–2.48); p = 0.006 | 1.69 (1.16–2.48); p = 0.007 | 1.78 (1.20–2.64); p = 0.004 | 1.72 (1.16–2.56); p = 0.005 |
| All cardiovascular events (all coronary events, nonfatal and fatal stroke, death from other atherosclerotic diseases) | 1.55 (1.15–2.10); p = 0.004 | 1.54 (1.14–2.10); p = 0.005 | 1.54 (1.12–2.11); p = 0.008 | 1.50 (1.09–2.07); p = 0.012 |
Model 1: adjustment for age, sex, ethnicity, hypertension/antihypertensive medications, systolic and diastolic blood pressure, pulse rate, smoking status and amount in pack-years, physical activity (intentional exercise), diabetes/antihyperglycemic treatment, total cholesterol, HDL, use of lipid-modifying medications, and waist-to-hip ratio. For heart failure, interim MI was additionally included and modeled as a time-dependent covariate.
Model 2: adjustment for covariates in model 1 and biomarkers of reported association with myocardial dysfunction and/or cardiovascular events in asymptomatic populations (fasting glucose, serum creatinine, C-reactive protein, fibrinogen, interleukin 6, and homocysteine).
Model 3: adjustment for covariates in model 2 and global LV systolic dysfunction (LV ejection fraction <61%, 10th percentile of reference MESA population).
Model 4: adjustment for covariates in model 3 and LV hypertrophy (LV mass index >90th sex-specific percentile of reference MESA population).