TABLE 3.
MULTIVARIABLE MODELS OF GLOBAL INITIATIVE FOR OBSTRUCTIVE LUNG DISEASE–UNCLASSIFIED STATUS
| OR | 95% CI | |
| GOLD-U versus Smoking Control Subjects | ||
| Clinical variables* | ||
| Resting O2 saturation | 0.867 | 0.803–0.935 |
| BMI | 1.041 | 1.016–1.065 |
| History CHF | 5.016 | 1.877–13.409 |
| 6MWD (per 100 ft) | 0.890 | 0.857–0.924 |
| Clinical + radiographic variables† | ||
| Diabetes mellitus | 1.579 | 1.016–2.452 |
| Pack-years | 1.010 | 1.003–1.018 |
| Resting O2 saturation | 0.882 | 0.809–0.960 |
| BMI | 1.030 | 1.004–1.057 |
| 6MWD (per 100 ft) | 0.948 | 0.905–0.993 |
| History of CHF | 3.460 | 1.107–10.817 |
| TLC,‡ L | 0.767 | 0.651–0.904 |
| Subsegmental wall area % | 1.302 | 1.202–1.409 |
| GOLD-U versus subjects with COPD (GOLD stages 2−4) | ||
| Clinical variables§ | ||
| Age (per 10 yr) | 0.630 | 0.490–0.810 |
| BMI | 1.121 | 1.089–1.154 |
| Resting O2 saturation | 1.244 | 1.149–1.347 |
| Bronchodilator response | 0.266 | 0.167–0.424 |
| 6MWD (per 100 ft) | 1.098 | 1.047–1.150 |
| Asthma (current) | 0.407 | 0.235–0.705 |
| Nonwhite race | 1.458 | 0.926–2.295‖ |
| Diabetes mellitus | 2.317 | 1.427–3.761 |
| Chronic bronchitis | 0.533 | 0.339–0.838 |
| Clinical + radiographic variables¶ | ||
| Bronchodilator response | 0.321 | 0.180–0.572 |
| Resting O2 saturation | 1.159 | 1.048–1.283 |
| BMI | 1.066 | 1.028–1.105 |
| % Emphysema | 0.752 | 0.661–0.856 |
| % Gas trapping | 0.926 | 0.899–0.954 |
| TLC,‡ L | 0.670 | 0.527–0.851 |
| Subsegmental wall area % | 0.774 | 0.693–0.865 |
Definition of abbreviations: BMI = body mass index; CHF = congestive heart failure; CI = confidence interval; COPD = chronic obstructive pulmonary disease; GOLD-U = Global Initiative for Obstructive Lung Disease–Unclassified; TLC = total lung capacity; 6MWD = 6-min walk distance.
Variables tested but not retained in the final model include: nonwhite race, pack-years, chronic cough, physician-diagnosed asthma (ever), resting heart rate, history of diabetes, hypertension, transient ischemic attack, and obstructive sleep apnea.
Final model includes race as nonsignificant confounder. Additional variables tested but not retained in the final model include: chronic cough, physician-diagnosed asthma (ever), resting heart rate, history of hypertension, transient ischemic attack, obstructive sleep apnea, and % emphysema.
TLC based on chest computed tomography measurements
Final model includes current smoking and pack-years as nonsignificant confounders. No additional variables were tested.
Becomes nonsignificant after addition of confounders.
Final model includes race as a nonsignificant confounder. Additional variables tested but not retained in the final model include: age, diabetes, 6MWD, current smoking, pack-years, chronic bronchitis, and physician-diagnosed asthma (current).