Table 3.

Association of potential explanatory variables with a greater difference between lung age and chronological age.

	Vitalograph standard equation		Equation based on present data
	P-value	Change in lung age ± SE	P-value	Change in lung age ± SE
Dyspnea	<0.0001	6.5 ± 0.32	<0.0001	5.8 ± 0.47
Smoker	<0.0001	5.7 ± 0.37	<0.0001	6.3 ± 0.53
COPD	<0.0001	13.9 ± 0.94	<0.0001	14.1 ± 1.30
Asthma	<0.0001	8.3 ± 0.64	<0.0001	8.8 ± 0.91
Chronological age decile	<0.0001	See Figure 2	<0.0001	See Figure 2
Height decile	<0.0001	See Figure 2	<0.0001	See Figure 2
Cough	<0.0001	1.7 ± 0.34	<0.0001	2.0 ± 0.49
Other airway disease	<0.0001	3.2 ± 0.69	0.0021	3.1 ± 1.00
Common cold	0.0006	1.5 ± 0.45	0.0048	1.8 ± 0.65
Patient-reported allergies	0.0052	−0.9 ± 0.33	0.21	−0.7 ± 0.46

Analyses were performed based on the equations used by the Vitalograph (“Vitalograph standard equation”) and the equation we have empirically derived in the subset of non-smokers not reporting any airway disease in this study (“equation based on present data”). Results are based on a selection procedure to identify predictors of high lung age using a multivariate linear model and shown as parameter estimates with SE and corresponding p-value. Parameter estimates are displayed in the sequence of inclusion into the model (based on F-tests). For categorical variables effects are relative to the respective control group of non-smoker, and absence of cough, allergies, COPD, asthma, dyspnea, or other airway disease. Note that the parameter estimates for chronological age and height are given for gender-specific deciles, respectively. Size of home town is not included because it did not reach the predefined threshold level of p < 0.15 for staying in the model. The parameter estimate for change in lung age is expressed in years.