Adjusted odds ratios and 95% confidence intervals
Variable | Contrast | Odds ratio (95% confidence interval) | Test statistic |
---|---|---|---|
IVSDT (mm) | 12:10 | 1.13 (0.82–1.56) | χ22 = 3.48, P = 0.176 |
LVESD (mm) | 35:28 | 1.38 (0.95–2.01) | χ12 = 2.81, P = 0.094 |
LVEF (%) | - | - | χ22 = 8.02, P = 0.018 |
- | 46:61 | 0.34 (0.16–0.73) | - |
- | 69:61 | 1.07 (0.67–1.71) | - |
LVPWDT (mm) | 12:9 | 1.35 (0.83–2.20) | χ12 = 1.43, P = 0.232 |
Aortic regurgitation | Mild to severe: none or trivial | 0.71 (0.39–1.28) | χ12 = 1.32, P = 0.250 |
Mitral regurgitation | Mild to severe: none or trivial | 0.96 (0.58–1.58) | χ12 = 0.03, P = 0.868 |
LVOT velocity (m/s) | - | - | χ22 = 8.20, P = 0.017 |
- | 0.9:1.1 | 0.51 (0.32–0.82) | - |
- | 1.4:1.1 | 1.20 (0.75–1.91) | - |
LV cardiac index [L/(min·m2)] | 3.6:2.6 | 1.05 (0.78–1.43) | χ12 = 0.11, P = 0.742 |
MV E:e’ | 16.7:8.6 | 1.42 (1.00–2.02) | χ12 = 3.85, P = 0.050 |
AVSP velocity (m/s) | 1.7:1.2 | 1.08 (0.84–1.40) | χ12 = 0.37, P = 0.543 |
For all continuous variables except LVOT and LVEF, the odds ratios represent an interquartile increase in that variable (i.e., an increase from the 25th to the 75th percentile). For LVOT and LVEF, which appear to be non-monotonically related to the outcome (see Figure 3), odds ratios are calculated at 2 data points compared with the median (i.e., 50th percentile) to represent the effects of having an extremely low value (10th percentile) and an extremely high value (90th percentile). As IVSDT, LVOT, and LVEF each showed evidence of nonlinearity, these variables were modeled with regression splines, and their overall tests of the association are based on 2 degrees of freedom. -: no data. IVSDT: interventricular septum diastolic thickness; LVOT: left ventricular outflow tract; LV: left ventricular; LVESD: left ventricular end-systolic dimension; LVPWDT: left ventricular posterior wall diastolic thickness; MV: mitral valve; AVSP: aortic valve systolic peak; LVEF: left ventricular ejection fraction