Table 2. Description of the 527 COPD patients based on phenotypes identified by cluster analysis.
| Phenotype 1 | Phenotype 2 | Phenotype 3 | P value | |
| n = 219 | n = 99 | n = 209 | ||
| DATA USED IN THE CLUSTER ANALYSIS | ||||
| Quantitative data | ||||
| Age, yrs. | 62 [58–68] | 61 [57–66] | 72 [65–77] | <0.001 |
| BMI, kg/m2 | 25 [23–28] | 20 [18–22] | 26 [24–29] | <0.001 |
| FEV1, % predicted | 80 [65–94] | 29 [21–37] | 44 [36–58] | <0.001 |
| Dyspnoea, mMRC scale | 0 [0–1] | 2 [1]–[3] | 2 [2]–[3] | <0.001 |
| Clinical COPD Questionnaire, Total | 1.8 [1.0–3.0] | 6.8 [5.0–9.0] | 6.3 [4.5–8.0] | <0.001 |
| TGV, % predicted | 126 [112–147] | 195 [177–220] | 139 [115–161] | <0.001 |
| DLCO, % predicted | 74 [60–86] | 34 [25–39] | 47 [38–61] | <0.001 |
| Categorical data | ||||
| CT scan * | ||||
| Emphysema present, % | 48 | 96 | 81 | <0.001 |
| Alveolar destruction | ||||
| Absent (%) | 52 | 4 | 19 | <0.001 |
| Mild (%) | 35 | 9 | 32 | |
| Moderate (%) | 8 | 43 | 27 | |
| Severe (%) | 5 | 44 | 22 | |
| Bronchial thickening | ||||
| Mild (%) | 57 | 36 | 20 | <0.001 |
| Moderate (%) | 33 | 45 | 52 | |
| Severe (%) | 9 | 19 | 27 | |
| Bronchiectasis, % | 14 | 31 | 33 | <0.001 |
| Comorbidities | ||||
| Ischemic heart disease, % | 14 | 17 | 34 | <0.001 |
| Stroke, % | 0 | 2 | 8 | <0.001 |
| Peripheral arterial disease * , % | 10 | 11 | 19 | 0.08 |
| Diabetes, % | 9 | 11 | 20 | 0.003 |
| Muscle weakness * , % | 18 | 47 | 42 | <0.001 |
| Osteoporosis, % | 8 | 31 | 22 | <0.001 |
| Anaemia, % | 5 | 15 | 10 | 0.005 |
| DATA NOT USED IN THE CLUSTER ANALYSIS | ||||
| Male sex, % | 80 | 66 | 80 | 0.007 |
| GOLD stage | <0.001 | |||
| G I, % (% of all GOLD stage I) | 50 (91) | 0 (0) | 5 (9) | |
| G II,% (% of all GOLD stage II) | 39 (51) | 8 (5) | 36 (44) | |
| G III, % (% of all GOLD stage III) | 9 (13) | 37 (25) | 44 (62) | |
| G IV, % (% of all GOLD stage IV) | 2 (5) | 55 (61) | 15 (35) | |
| Source of recruitment | ||||
| NELSON cohort, % (% NELSON) | 67 (95) | 0 (0) | 3 (5) | <0.001 |
| LEUVEN clinic, % (% LEUVEN) | 33 (19) | 100 (27) | 97 (54) | <0.001 |
| Smoking, pack-year | 44 [34–60] | 48 [33–56] | 48 [31–65] | 0.89 |
| FEV1, L | 2.4 [1.9–2.9] | 0.8 [0.6–1.0] | 1.25 [1.0–1.6] | <0.001 |
| FVC, % predicted | 105 [92–118] | 72 [61–86] | 82 [69–94] | <0.001 |
| FVC, L | 4.0 [3.3–4.6] | 2.6 [1.9–3.2] | 2.8 [2.3–3.3] | <0.001 |
| FEV1/FVC ratio | 0.62 [0.53–0.67] | 0.33 [0.27–0.38] | 0.44 [0.37–0.55] | <0.001 |
| RV,% predicted | 128 [109–149] | 227 [186–268] | 144 [113–174] | <0.001 |
| TLC, % predicted | 106 100–118] | 128 [116–139] | 104 [91–114] | <0.001 |
| Raw, % predicted | 174 [140–223] | 274 [215–379] | 249 [198–317] | <0.001 |
| Sgaw, % predicted | 68 [51–87] | 29 [22–40] | 44 [32–64] | <0.001 |
| Kco, % predicted | 86 [71–98] | 53 [45–62] | 70 [56–91] | <0.001 |
| Mortality, n (%) | 1 (0.5%) | 20 (20.6%) | 29 (14.3%) | <0.001 |
*
% missing data: Phenotype 1∶67%; Phenotype 2∶1%, Phenotype 3∶4%.
P values correspond to comparisons between the 3 phenotypes using Kruskal-Wallis or Chi-square tests, as appropriate.