Cox proportional hazards models for prediction of CV disease mortality

Total population (n = 994)
No CHD (n = 742)
Possible or definite CHD (n = 252)
HR95% CIχ2PHR95% CIχ2PHR95% CIχ2P
Adjusted for study center only
  ≥50 μV9.545.43–16.7361.9<0.00112.034.95–29.2530.1<0.0016.152.76–13.7119.7<0.001
  PCA ratio*3.171.81–5.5616.3<0.0012.380.98–5.793.70.0552.811.26–
  QTc >460 ms2.071.13–3.795.60.0182.221.03–4.943.90.0481.720.68–4.331.30.251
Adjusted for multiple covariates
   ≥50 μV3.681.70–7.9610.9<0.0017.892.28–27.2810.7<0.0016.842.90–16.1419.3<0.001
  PCA ratio*2.611.33––6.471.80.1862.240.91–5.513.10.079
  QTc >460 ms1.870.92–3.793.00.0822.551.01–6.563.80.0492.871.07–7.684.40.036
  • *

    * >32.0% in women, >24.6% in men.

  • Cox model with all three ECG variables entered, adjusted for age, BMI, diastolic and systolic blood pressures, fasting glucose, GHb, HDL and LDL cholesterol levels, triglyceride level, albuminuria, alcohol use, history of smoking, QTc interval, and study center. Prevalent CHD was included as a covariate only in the total population. HR, hazard ratio.