Table 2– Pooled estimates of effects of risk factors on stage I or higher chronic obstructive pulmonary disease (COPD),^# and on percentage forced expiratory volume in 1 s (FEV₁)/forced vital capacity (FVC) ratio, adjusted for all risk factors in the regression model, including pack-years of smoking

	Stage I or higher COPD				% FEV₁/FVC
	OR (95% CI)	p-value	I² %	p-value for heterogeneity	Regression coefficient^¶ (95% CI)	p-value	I² %	p-value for heterogeneity
Females	1.10 (0.85–1.43)	0.46	57	0.004	1.16 (0.70–1.57)	<0.001	39	0.069
Age per 10-yr difference	1.52 (1.35–1.71)	<0.001	60	0.002	-2.69 (-2.88– -2.49)	<0.001	26	0.17
BMI kg·m⁻²
<25	1.00				0.00
25–30	0.58 (0.46– 0.72)	<0.001	43	0.045	1.87 (1.29–2.45)	<0.001	54	0.008
>30	0.50 (0.37–0.67)	<0.001	50	0.020	2.54 (1.87–3.20)	<0.001	50	0.017
Education⁺	0.76 (0.67–0.87)	<0.001	26	0.18	0.54 (0.29–0.80)	<0.001	0	0.99
Hospitalisation for breathing problems before age 10 yrs	2.35 (1.42– 3.91)	0.001	54	0.011	-1.35 (-2.44–-0.27)	0.014	32	0.12
Current smoking	1.31 (1.00–1.72)	0.049	52	0.012	-0.38 (-1.10–0.34)	0.30	55	0.007
Passive smoking^§	1.24 (1.05–1.47)	0.012	0	0.45	-0.61 (-1.06– -0.15)	0.009	14	0.30
Doctor ever-diagnosed tuberculosis	1.78 (1.17–2.72)	0.007	41	0.070	-1.98 (-4.23–0.26)	0.083	80	<0.001
Family history of COPD^ƒ	1.50 (1.19– 1.90)	0.001	31	0.13	-0.95 (-1.51– -0.39)	0.001	26	0.18
Years working in dusty jobs per 10-yr difference	1.04 (0.98– 1.11)	0.17	0	0.71	-0.33 (-0.50– -0.15)	<0.001	0	0.49
Regular exposure to dust in present job	0.86 (0.61– 1.21)	0.40	36	0.091	0.61 (-0.02–1.24)	0.058	28	0.16
Regular exposure to fumes in present job	0.91 (0.67–1.24)	0.55	0	0.97	0.37 (-0.22–0.97)	0.22	7	0.38
Years exposed to biomass fires per 10-yr difference^##
For heating	1.01 (0.94–1.10)	0.72	0	0.63	0.04 (-0.17–0.26)	0.71	0	0.47
For cooking^¶¶	1.02 (0.68–1.54)	0.91	2	0.42	-0.20 (-2.16–1.76)	0.84	32	0.15

BMI: body mass index. See online supplementary material for more details of model fitting and coding of risk factors. The model adjusts for a quadratic effect of pack-years of smoking, which is not shown in the table because the separate coefficients of pack-years and pack-years squared are difficult to interpret. Table 1 shows how pack-years of smoking affected prevalence of COPD at different sites. ^#: stage I or higher COPD defined as having an FEV₁/FVC ratio below the lower limit of normal given age and sex, using reference equations from National Health and Nutrition Examination Survey. ^¶: a negative regression coefficient indicates a reduction in FEV₁/FVC, i.e. poorer lung function. ⁺: effect per group, assuming a linear effect over the four groups of highest level of education, as follows: none; primary (primary or middle school); secondary (secondary school); tertiary (technical/vocational college or university). ^§: somebody else smoking in the subject's home in the last 2 weeks. ^ƒ: doctor ever-diagnosed mother, father, sister or brother with emphysema, chronic bronchitis or COPD. ^##: fires burning coal, coke, peat, wood, crop residue or dung. ^¶¶: years of equivalent continuous exposure are calculated by multiplying number of years over which subject was exposed by number of hours of exposure per day divided by 24.