Multivariable analyses of predictors of exacerbations in the past 1 year shown as crude and adjusted ORs, in individuals with and without COPD (results shown were from full predictor model*)
| Exacerbation in the past 1 year | ||||
|---|---|---|---|---|
| Non-COPD (post FEV1/FVC≥0.7) and no DDX asthma/emphysema/chronic bronchitis/COPD | COPD (post FEV1/FVC<0.7) | |||
| Crude ORs (95% CI) | Adjusted ORs† (95% CI) | Crude ORs (95% CI) | Adjusted ORs† (95% CI) | |
| Age categories | ||||
| 40–49 | 1.00 | 1.00 | 1.00 | 1.00 |
| 50–59 | 0.75 (0.50 to 1.14) | 0.65 (0.41 to 1.05) | 1.05 (0.45 to 2.48) | 1.17 (0.40 to 3.44) |
| 60–69 | 0.53 (0.31 to 0.89)* | 0.55 (0.30 to 1.01) | 0.85 (0.36 to 1.99) | 0.69 (0.23 to 2.09) |
| 70 and above | 0.78 (0.44 to 1.36) | 0.74 (0.37 to 1.48) | 0.64 (0.27 to 1.52) | 0.52 (0.16 to 1.71) |
| Women (men as ref.) | 1.53 (1.06 to 2.21)* | 1.67 (1.08 to 2.56)* | 1.37 (0.84 to 2.26) | 1.14 (0.61 to 2.13) |
| BMI categories | ||||
| <20 | 1.57 (0.65 to 3.80) | 1.59 (0.58 to 1.48) | 0.63 (0.08 to 4.99) | 0.79 (0.08 to 7.48) |
| (20, 25) | 1.00 | 1.00 | 1.00 | 1.00 |
| (25, 30) | 0.97 (0.62 to 1.51) | 0.91 (0.56 to 1.48) | 1.25 (0.65 to 2.42) | 1.07 (0.49 to 2.35) |
| 30 and above | 1.30 (0.83 to 2.03) | 0.79 (0.47 to 1.33) | 2.32 (1.20 to 4.46)* | 0.92 (0.39 to 2.15) |
| Ever smoking of cigarettes (never as ref.) | 1.67 (1.17 to 2.40)* | 1.28 (0.85 to 1.93) | 1.19 (0.69 to 2.07) | 0.58 (0.29 to 1.15) |
| Pack years of cigarettes | ||||
| 0 | 1.00 | ‡ | 1.00 | ‡ |
| (0, 10) | 2.10 (1.36 to 3.26)* | ‡ | 0.16 (0.02 to 1.24) | ‡ |
| (10, 20) | 1.54 (0.88 to 2.70) | ‡ | 0.85 (0.33 to 2.20) | ‡ |
| 20 and above | 1.41 (0.87 to 2.26) | ‡ | 1.57 (0.89 to 2.75) | ‡ |
| Passive smoking (yes; no as ref.) | 1.04 (0.55 to 1.95) | ‡ | 0.93 (0.45 to 1.93) | ‡ |
| Chronic cough (yes; no as ref.) | 2.47 (1.55 to 3.95)* | 1.14 (0.63 to 2.06) | 4.18 (2.52 to 6.92)* | 1.34 (0.65 to 2.77) |
| Chronic phlegm (yes; no as ref.) | 2.74 (1.63 to 4.61)* | 1.82 (0.94 to 3.49) | 3.24 (1.93 to 5.42)* | 1.01 (0.47 to 2.15) |
| Wheezing (yes; no as ref.) | 8.45 (5.84 to 12.22)* | 6.69 (4.39 to 10.21)* | 5.98 (3.22 to 11.12)* | 2.70 (1.25 to 5.81)* |
| Breathlessness (yes; no as ref.) | 2.25 (1.52 to 3.31)* | 1.35 (0.86 to 2.13) | 3.71 (2.07 to 6.64)* | 1.89 (0.90 to 3.96) |
| Self-reported doctor diagnosis of asthma (yes; no as ref.) | § | § | 2.39 (1.45 to 3.94)* | 0.70 (0.35 to 1.41) |
| Self-reported doctor diagnosis of emphysema/chronic bronchitis/COPD (yes; no as ref.) | § | § | 4.03 (2.42 to 6.69)* | 1.65 (0.80 to 3.41) |
| Comorbidities¶ (yes; no as ref.) | 1.20 (0.83 to 1.73) | 1.25 (0.78 to 2.01) | 1.20 (0.73 to 1.96) | 1.04 (0.54 to 1.99) |
| Childhood hospitalisation for breathing problem (yes; no as ref.) | 1.60 (0.77 to 3.34) | 1.42 (0.61 to 3.30) | 1.68 (0.80 to 3.55) | 0.85 (0.29 to 2.48) |
| Use of respiratory medications (yes; no as ref.) | 2.82 (1.98 to 4.02)* | 1.96 (1.31 to 2.94)* | 4.65 (2.64 to 8.20)* | 2.29 (1.08 to 4.84)* |
| Self-perceived health status Fair or poor (excellent/very good/good as ref.) | 3.73 (2.40 to 5.78)* | 2.26 (1.27 to 4.02)* | 5.76 (3.40 to 9.76)* | 2.42 (1.19 to 4.92)* |
| Post-BD FEV1, % predicted | ||||
| >80 | 1.00 | 1.00 | 1.00 | 1.00 |
| (50, 80) | 1.20 (0.67 to 2.16) | 0.60 (0.29 to 1.23) | 2.52 (1.41 to 4.49)* | 1.23 (0.60 to 2.51) |
| <50 | § | § | 9.23 (4.48 to 18.98)* | 2.79 (1.04 to 7.53)* |
| %Δ FVCi>12% (≤12% as ref.)** | 0.49 (0.07 to 3.55) | 0.29 (0.03 to 2.66) | 1.11 (0.59 to 2.09) | 0.62 (0.25 to 1.56) |
| %Δ FEV1i>12% (≤12% as ref.)** | 1.13 (0.45 to 2.82) | 1.10 (0.40 to 3.02) | 1.62 (0.93 to 2.83) | 1.13 (0.51 to 2.51) |
*p<0.05.
†Two predictive models were explored: a parsimonious model (unadjusted predictors that were significant at p<0.05), and the fuller model that included additional predictors that were thought a priori to be associated with a risk of exacerbation if p values were <0.25. The adjusted ORs (95% CIs) shown were from fuller model as results from both models were similar.
‡Pack years is not included in the multiple logistic model to avoid the colinearity with ever smoking. Passive smoking is not included in the multiple logistic models because the crude ORs have a non-significant p value>0.25.
§No data in this category.
¶Comorbidities include heart disease, hypertension, diabetes, and stroke.
**%Δ FVCi=(post FVC – pre FVC)/pre FVC, %Δ FEV1i=(post FEV1 – pre FEV1)/pre FEV1.
BD, bronchodilator; BMI, body mass index.