Table 2. Cox model HR and 95% CI for lung cancer mortality in lifetime smoky coal users in Xuanwei, China, 1976–1996a.
Men and women
|
Men
|
Women
|
||||
---|---|---|---|---|---|---|
HR (95% CI) | P-value | HR (95% CI) | P-value | HR (95% CI) | P-value | |
No stove change | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
Stove change | 0.52 (0.41–0.64) | <0.0001 | 0.62 (0.46–0.82) | 0.001 | 0.41 (0.29–0.57) | <0.0001 |
Used annual average <2 tons of smoky coal | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
Used annual average ⩾2 to <3 tons of fuel | 1.61 (1.27–2.03) | <0.0001 | 1.27 (0.93–1.73) | 0.133 | 2.07 (1.43–3.00) | 0.0001 |
Used annual average ⩾3 to <4 tons of fuel | 2.19 (1.73–2.79) | <0.0001 | 1.77 (1.29–2.43) | 0.0004 | 2.70 (1.85–3.93) | <0.0001 |
Used annual average ⩾4 tons of fuel | 2.37 (1.84–3.06) | <0.0001 | 1.69 (1.21–2.38) | 0.002 | 3.32 (2.24–4.92) | <0.0001 |
No lung cancer in spouse | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
Lung cancer in spouse | 1.77 (1.43–2.18) | <0.0001 | 1.47 (1.07–2.02) | 0.019 | 2.12 (1.60–2.82) | <0.0001 |
No lung cancer in parents, siblings, and children | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
Lung cancer in parents, siblings, and children | 1.36 (1.08–1.70) | 0.008 | 1.48 (1.10–1.99) | 0.010 | 1.20 (0.85–1.69) | 0.311 |
No history of COPD | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
History of COPD | 1.24 (1.08–1.43) | 0.003 | 1.33 (1.10–1.62) | 0.004 | 1.19 (0.97–1.46) | 0.103 |
No history of tuberculosis | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
History of tuberculosis | 1.38 (0.80–2.39) | 0.252 | 1.95 (0.96–3.95) | 0.065 | 0.89 (0.37–2.15) | 0.789 |
<7 waking hours indoors per day, until 20 years old | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
⩾7 waking hours indoors per day, until 20 years old | 1.11 (0.98–1.26) | 0.089 | 1.03 (0.86–1.23) | 0.747 | 1.21 (1.02–1.45) | 0.034 |
Average of <2 rooms in all houses throughout life | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
Average of ⩾2 rooms in all houses throughout life | 1.10 (0.97–1.25) | 0.155 | 1.18 (0.98–1.42) | 0.075 | 1.03 (0.85–1.24) | 0.755 |
Illiterate | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | |||
Literate | 1.19 (1.02–1.38) | 0.023 | 1.15 (0.95–1.39) | 0.146 | 1.27 (0.97–1.65) | 0.082 |
Never coal miner | 1.00 (reference) | 1.00 (reference) | Not applicable | |||
Ever coal miner | 3.34 (2.56–4.36) | <0.0001 | 3.39 (2.57–4.47) | <0.0001 | Not applicable | |
<20 years of smoking | 1.00 (reference) | 1.00 (reference) | Not applicable | |||
⩾20 years of pipe smoking | 0.86 (0.69–1.08) | 0.197 | 0.84 (0.64–1.11) | 0.223 | Not applicable | |
20–40 years of cigarette smoking | 1.21 (0.96–1.53) | 0.114 | 1.22 (0.91–1.63) | 0.179 | Not applicable | |
⩾40 years of cigarette smoking | 1.42 (1.04–1.94) | 0.027 | 1.44 (0.99–2.09) | 0.054 | Not applicable | |
Never cooked | Not applicable | 1.00 (reference) | Not applicable | |||
Ever cooked | Not applicable | 0.94 (0.73–1.22) | 0.650 | Not applicable | ||
<20 years of cooking | 1.00 (reference) | Not applicable | 1.00 (reference) | |||
20–40 years of cooking | 1.27 (1.03–1.55) | 0.023 | Not applicable | 1.45 (0.91–2.32) | 0.121 | |
⩾40 years of cooking | 1.44 (1.12–1.86) | 0.004 | Not applicable | 1.53 (0.92–2.54) | 0.103 |
CIs=confidence intervals; HRs=hazard ratios.
Cox models include the time-dependent variables for change to portable stove, average tons of fuel used annually (2–<3 tons, 3–<4 tons, ⩾4 tons, compared with <2 tons), years of smoking (20–40 years, ⩾40 years, compared with <20 years), and years of cooking (20–40 years, ⩾40 years, compared with <20 years), along with dummy variables for history of spousal lung cancer (yes vs no), family history of lung cancer (parents, siblings, or children) (yes vs no), history of chronic bronchitis (yes vs no), history of tuberculosis (yes vs no), number of hours spent indoors until the age of 20 years (<7 vs ⩾7 h), average number of rooms in all households throughout life (<2 rooms vs ⩾2 rooms), educational status (illiterate vs literate), and history of working as a miner (yes vs no). Sex-specific Cox-regression models include the same covariates, as well as, history of ever cooking (ever vs never) in men. Birth cohort effects are controlled by stratification within the model by 5-year intervals (born during 1917–1921, 1922–1926, 1927–1931, 1932–1936, 1937–1941, 1942–1946, and 1947–1951).