| Variables | Female | Male | ||
| COPD | No COPD | COPD | No COPD | |
| Subjects | 362 | 2850 | 397 | 1705 |
| Age yrs | 64.8±0.7# | 55.4±0.2¶¶ | 63.5±0.6# | 54.3±0.3 |
| BMI kg·m−2 | 27.2±0.3#,¶ | 28.8±0.1++ | 26.5±0.2+ | 27.4±0.1 |
| BMI categories | ||||
| Underweight | 29 (8.0)#,¶ | 79 (2.8)++ | 22 (5.5)+ | 56 (3.3) |
| Normal | 100 (27.6) | 693 (24.3) | 130 (32.8) | 423 (24.8) |
| Overweight | 135 (37.3) | 1066 (37.4) | 168 (42.3) | 807 (47.3) |
| Obese | 98 (27.1) | 1012 (35.5) | 77 (19.4) | 419 (24.6) |
| White subjects | 236 (65.2)+ | 1526 (53.8)§§ | 253 (63.7) | 863 (50.8) |
| Education yrs | 6.5±0.3+ | 7.5±0.1++ | 6.8±0.3# | 8.1±0.2 |
| Employed | 112 (30.9)#,§ | 1303 (45.7)++ | 205 (51.6)# | 1299 (76.2) |
| Smoking exposure pack-yrs | 11.6±1.2#,§ | 6.0±0.3++ | 26.5±1.4# | 14.1±0.5 |
| Smoking status | ||||
| Current | 113 (31.2)ƒ,§ | 706 (24.8)++ | 160 (40.3)+ | 604 (35.5) |
| Former | 75 (20.7) | 603 (21.2) | 172 (43.3) | 617 (36.3) |
| Never | 174 (48.1) | 1541 (54.1) | 65 (16.4) | 481 (28.3) |
| Cough | 118 (32.6)# | 582 (20.4)§§ | 120 (30.2)# | 288 (16.9) |
| Phlegm | 97 (26.8)# | 471 (16.5) | 118 (29.7)# | 308 (18,1) |
| Wheezing | 142 (39.2)# | 620 (21.8) | 153 (38.5)# | 353 (20.7) |
| Dyspnoea | 214 (60.5)##,§ | 1469 (52.0)++ | 165 (41.9)## | 561 (33.3) |
| Self-reported tuberculosis | 19 (5.3)+ | 66 (2.3)§§ | 20 (5.0)# | 22 (1.3) |
| Comorbidity score | 1.35±0.05#,§ | 1.10±0.02++ | 1.01±0.05+ | 0.79±0.02 |
| Any respiratory medication | 69 (19.1)#,¶ | 172 (6.0)++ | 44 (11.1)# | 55 (3.2) |
| Any bronchodilator | 64 (17.7)#,¶ | 158 (5.5)++ | 43 (10.8)# | 47 (2.8) |
| Any corticosteroid | 25 (6.9)# | 45 (1.6)§§ | 17 (4.3)# | 15 (0.9) |
| Prior spirometry | 68 (18.8)# | 239 (8.4)++ | 84 (21.2)# | 212 (12.5) |
Data are presented as n, mean±se and n (%). Statistical tests used were as follows: for nominal variables, Pearson Chi-squared adjusted for survey design; for ordinal variables, Mann–Whitney test and ordered logistic regression adjusted for survey design (p-values reflect Mann–Whitney test); for continuous variables, Wald test adjusted for survey design. BMI: body mass index. #: p<0.0001 (COPD versus no COPD); ¶: p<0.05 (COPD female versus COPD male); +: p<0.001 (COPD versus no COPD); §: p<0.001 (COPD female versus COPD male); ƒ: p<0.05 (COPD versus no COPD); ##: p<0.01 (COPD versus no COPD); ¶¶: p<0.001 (no COPD female versus no COPD male); ++: p<0.0001 (no COPD female versus no COPD male); §§: p<0.05 (no COPD female versus no COPD male).