Table 3. Risk of lung cancer death by cumulative silica exposure in a linear spline model with one knot at 10 mg m− 3-years, 1946–2003.
| Model | β1a 95% CI | β2b 95% CI | Deviance | RRc 95% CI 5 mg m−3-years | RR 95% CI 15 mg m − 3-years |
|---|---|---|---|---|---|
| Unadjusted | 0.067 (0.039; 0.094) | 0.088 (0.058; 0.119) | 30 768.1 | 1.34 (1.20; 1.47) | 2.44 (1.96; 2.61) |
| Adjusted for radon d | 0.015 (−0.008; 0.038) | 0.072 (0.042; 0.102) | 30 692.1 | 1.08 (0.96; 1.19) | 1.59 (1.09; 2.08) |
| Adjusted for radon and arsenic e | 0.012 (−0.011; 0.035) | 0.068 (0.038; 0.098) | 30 688.2 | 1.06 (0.95; 1.18) | 1.52 (1.24; 1.79) |
| Adjusted for radon (with exponential effect modifiers f) and arsenic | −0.006 (−0.028; 0.015) 0.000 (fixedg) | 0.066 (0.038; 0.094) 0.061 (0.039; 0.083) | 30 604.2 30 604.5 | 0.97 (0.86; 1.08) 1.00 (fixed) | 1.24 (0.98; 1.49) 1.31 (1.20; 1.42) |
| 0-year lag 10-year lag 15-year lag | 0.000 (fixed) 0.000 (fixed) 0.000 (fixed) | 0.061 (0.039; 0.083) 0.060 (0.037; 0.082) 0.062 (0.039; 0.085) | not comparable | 1.00 (fixed) 1.00 (fixed) 1.00 (fixed) | 1.31 (1.20;1.42) 1.30 (1.19; 1.41) 1.31 (1.20; 1.43) |
| Silica with exponential effect modifier attained age (centred at 64 years) adjusted for radon (with exponential modifiers h) and arsenic | 0.000 (fixed) 0.002 (−0.014; 0.018) | 0.073 (0.046; 0.100) 0.073 (0.043; 0.100) | 30 592.5 30 592.3 | 1.00 (fixed) 1.01 (0.93; 1.09) | 1.37 (1.23; 1.50) 1.39 (1.16; 1.63) |
| 50 years attained age | 0.007 (−0.040; 0.057) | 0.223 (0.142; 0.322) | 30 592.3 | 1.04 (0.80; 1.29) | 2.21 (1.43; 3.02) |
| 60 years | 0.003 (−0.019; 0.025) | 0.105 (0.066; 0.134) | 30 592.3 | 1.15 (0.90; 1.13) | 1.55 (1.30; 1.84) |
| 70 years | 0.001 (0.008; 0.011) | 0.045 (0.021; 0.068) | 30 592.3 | 1.01 (0.96; 1.06) | 1.24 (1.07; 1.39) |
| 80 years | 0.001 (−0.004; 0.005) | 0.020 (0.004; 0.035) | 30 592.3 | 1.00 (0.98; 1.03) | 1.11 (1.02; 1.20) |
Abbreviations: CI=confidence interval; ERR=excess relative risk.
a,b
β1 and β2 describe the ERR in the following way, silk=10: 
c
Relative risk with 95% CI for a cumulative exposure of 5 mg m−3-years compared with 0.
d
Adjusted for cumulative radon exposure in a linear way.
e
Adjusted for cumulative radon and arsenic exposure in a linear way.
f
Exponential effect modifiers age at median exposure, time since median exposure and radon exposure rate.
g
Parameter β1 is fixed at 0.
h
Exponential effect modifiers, time since median exposure and radon exposure rate.