Table 5

Adjusted linear regression on cross-shift decline in FEV₁ (the residuals adjusted for age, gender, and height), and dust exposure

Exposure estimates	RES cross-shift FEV₁*
	+ smoking (n = 497)			− smoking (n = 639)			− smoking, − pine (n = 378)			− smoking, + pine (n = 260)
	Coeff.†	SE	p value	Coeff.†	SE	p value	Coeff.†	SE	p value	Coeff.†	SE	p value
Stratified by smoking and by use of pine. The estimates are labelled according to the method number.
n = number of persons. Numbers may vary due to missing values.
*Variables included in the final models: dust exposure, use of beech, body mass index.
†Unit: percentage decline in FEV₁/mg/m³ dust.
1. Individual, first round	−0.060	0.35	0.86	0.821	0.32	0.01	0.522	0.40	0.19	1.230	0.53	0.02
2. Individual, all rounds	0.022	0.37	0.95	0.792	0.34	0.02	0.343	0.41	0.42	1.474	0.57	0.01
3. Individual, 4 groups	−0.027	0.46	0.95	0.829	0.38	0.03	0.130	0.48	0.79	1.569	0.55	<0.01
4. 12 categories	−0.836	0.90	0.36	1.636	0.77	0.03	2.016	1.01	0.05	1.270	1.22	0.30
5. 5 categories	−0.143	0.60	0.81	0.296	0.53	0.57	0.046	0.66	0.94	0.799	0.88	0.36
6. Weighted estimate	−0.314	0.59	0.59	1.336	0.53	0.01	0.895	0.65	0.17	2.050	0.90	0.02
7. Mixed model 1	−0.479	0.59	0.42	0.801	0.57	0.16	0.248	0.71	0.73	1.631	0.93	0.08
8. Mixed Model 2	−0.247	0.62	0.69	1.248	0.56	0.03	0.696	0.71	0.33	2.020	0.92	0.03