TABLE 3. Clusters in Secondary Care.
| Cluster 1 |
Cluster 2 |
Cluster 3 |
Cluster 4 |
|||
|---|---|---|---|---|---|---|
| Variable | Secondary Care (n = 187) |
Early Onset, Atopic (n = 74) |
Obese, Noneosinophilic (n = 23) |
Early Symptom Predominant (n = 22) |
Inflammation Predominant (n = 68) |
Significance (P Value)* |
| Sex†, % female | 65.8 | 75.7 | 87 | 68.2 | 47.1 | <0.001 |
| Age, yr (SD) | 43.4 (15.9) | 39.4 (15.7) | 42.7 (11.1) | 35.5 (15.5) | 50.6 (15.1) | <0.001 |
| Age of onset†, yr (SD) | 20.3 (18.4) | 12.7 (12.9) | 15.4 (15.2) | 12.6 (15) | 32.6 (19.1) | <0.001 |
| Atopic status†, % positive | 73.8 | 83.8 | 65.2 | 81.8 | 63.2 | 0.024 |
| Body mass index†, kg/m2 (SD) | 28.5 (6.5) | 27.6 (4.5) | 40.9 (6.5) | 23.6 (3.1) | 27 (3.9) | <0.001 |
| Peak flow variability‡, amp % mean | 32.2 (0.48) | 46.1 (0.35) | 21.2 (0.76) | 24.2 (0.65) | 27.6 (0.36) | 0.002 |
| FEV1 change with bronchodilator‡, % | 12.8 (0.41) | 24.5 (0.31) | 9.3 (0.35) | 4.5 (0.33) | 9.8 (0.34) | <0.001 |
| Post-bronchodilator FEV1, % predicted (SD) | 82.1 (21.1) | 79.0 (21.9) | 79.0 (18.5) | 79.5 (26.1) | 87.2 (18.5) | 0.093 |
| Sputum eosinophil count†‡, % | 2.9 (0.99) | 4.2 (0.76) | 1.3 (1.01) | 0.1 (0.9) | 8.4 (0.64) | <0.001 |
| FeNO‡§, ppb | 43 (0.32) | 51.2 (0.36) | 24.2 (0.27) | 22.6 (0.30) | 53.1 (0.32) | <0.001 |
| Sputum neutrophil count, %‡ | 46.7 (0.32) | 45.4 (0.39) | 49.3 (0.22) | 51.3 (0.23) | 45.9 (0.29) | 0.892 |
| Modified JACS† (SD) | 2.02 (1.16) | 2.63 (0.93) | 2.37 (1.09) | 2.11 (1.11) | 1.21 (0.95) | <0.001 |
| Dose of inhaled corticosteroid, BDP equivalent/μg (SD) | 1,018 (539) | 1,168 (578) | 1,045 (590) | 809 (396) | 914 (479) | 0.008 |
| Long-acting bronchodilator use, % | 93.0 | 91.9 | 95.4 | 90.9 | 94.1 | 0.999 |
| Maintenance oral corticosteroid use, % | 31.7 | 32.4 | 22.7 | 22.7 | 36.8 | 0.604 |
| Median Nijmegen score (IQR) (% with score >23)∥ | 16 (7–26.5) | 20.5 (12–30.25) (44.6) | 23 (12–33) (52.2) | 16.5 (4.5–27.5) (31.8) | 9 (1–17) (19.1) | 0.004 |
| Median anxiety score (IQR) (% with score ≥11)∥ | 7 (4–10) | 7.5 (4.75–10.25) (24.3) | 8 (3–14) (34.8) | 6 (3.75–8.25) (13.6) | 6 (3–9) (19.1) | 0.34 |
| Median depression score (IQR) (% with score ≥11)∥ | 4 (2–7) | 4.5 (2–8) (13.5) | 5 (2–7) (4.3) | 4 (2–7) (4.5) | 3 (1–6) (7.4) | 0.104 |
| Courses of oral corticosteroids for asthma exacerbations, n/case/yr | 4.05 (2.33) | 4.62 (0.27) | 3.90 (0.38) | 3.57 (0.49) | 3.43 (0.27) | 0.02 |
| Hospital admissions for asthma, n/case/yr | 1.54 | 1.64 | 1.61 | 1.54 | 1.23 | 0.703 |
| Failed clinic appointments, % total appointments to DAC/yr | 20.0 | 26.2 | 15.7 | 19.0 | 14.8 | 0.027 |
Definition of abbreviations: amp = amplitude; BDP = beclomethasone diproprionate; DAC = difficult asthma clinic; IQR = interquartile range; JACS = Juniper Asthma Control Score; SD = standard deviation.
Anxiety and depression scores are obtained from the Hospital Anxiety and Depression Scale, a validated 14-point screening questionnaire. Scores of greater than 11 for either domain are suggestive of clinically important symptoms (25). Boldface type denotes population statistics. Columns headed “Cluster 3” and “Cluster 4” represent clusters not identified in the primary care asthma population.
Comparison between clusters using analysis of variance for continuous variables and χ2 test for proportions. As for the other tables, significance values for variables included in the cluster analysis are a product of the cluster algorithm and should not be further interpreted.
Variables included in the cluster analysis.
Geometric mean (log10 SD).
Measured with NIOX at a flow rate of 50 ml/second.
The Nijmegen score is obtained from responses to the Nijmegen questionnaire, a 16-point screening questionnaire for hyperventilation syndrome. Scores of greater than 23 are suggestive of clinically significant hyperventilation (24).