Table 4.
Multivariable modeling of the association of lung cancer risk with joint of measures of COPD.
| Measures | Nlung cancer cases | Ncontrols | Contrast | OR1 (95% CI) | p- value |
|---|---|---|---|---|---|
| Model 1 | |||||
| FEV1/FVC < 0.70/Emphysema by qCT-950insp (>4.8% threshold)2 | 131 | 454 | Ref (−/−) | -- | -- |
| 85 | 149 | +/− vs. −/− | 2.05 (1.44, 2.92) | <0.001 | |
| 12 | 29 | −/+ vs. −/− | 2.99 (1.39, 6.40) | 0.005 | |
| 52 | 81 | +/+ vs. −/− | 3.15 (1.96, 5.07) | <0.001 | |
| Model 2 | |||||
| FEV1/FVC < 0.70/Air trapping by qCT-856exp (>19.5% threshold)3 | 99 | 391 | Ref (−/−) | -- | -- |
| 49 | 99 | +/− vs. −/− | 2.08 (1.36, 3.20) | 0.001 | |
| 43 | 93 | −/+ vs. −/− | 2.42 (1.46, 3.99) | 0.001 | |
| 89 | 132 | +/+ vs. −/− | 3.95 (2.48, 6.28) | <0.001 | |
1
Odds ratios are adjusted for age, race, gender, pack years and total lung volume (inspiratory and expiratory).
2
Emphysema based on a threshold of 4.8% voxels < -950 HU in inspiratory scans across both lungs as described in COPDGene CT Workshop Group (24).
3
Air trapping based on a threshold of 19.5% voxels < -856 HU in expiratory scans across both lungs as described in COPDGene CT Workshop Group (24).