Evaluation of variation in exposure effect with age at exposure under three exposure weighting functions. Canadian nuclear industry workers study
| Exposure weight function | Residual deviance | Likelihood ratio statistic* | Degrees of freedom | p Value |
|---|---|---|---|---|
| *Test for variation in exposure effect with age at exposure. Model fit is compared to the model for total cumulative (that is, η = 0 for the step function; σ1 = 1 for the sigmoid function; and, φ1 = φ2 = 0 for the bilinear function with φ2 parameterised as a slope). Exposure assignment is lagged 10 years for analyses of lung cancer and cancers other than lung and leukaemia, and two years for analyses of leukaemia. Analyses of lung cancer, and cancers other than lung and leukaemia, were matched on attained age, sex, calendar year at risk, facility, socioeconomic status, duration of monitoring, and monitoring status; analyses of leukaemia were matched on attained age and adjusted for sex, calendar year at risk, facility, socioeconomic status, duration of monitoring, and monitoring status. | ||||
| Lung cancer | ||||
| Step function | 1234.12 | 3.86 | 1 | 0.05 |
| Sigmoid function | 1234.73 | 3.25 | 1 | 0.07 |
| Bilinear function | 1237.26 | 0.72 | 2 | 0.70 |
| Leukemia | ||||
| Step function | 79.67 | 3.82 | 1 | 0.05 |
| Sigmoid function | 79.93 | 3.56 | 1 | 0.06 |
| Bilinear function | 82.64 | 0.85 | 2 | 0.65 |
| Cancers other than lung and leukemia | ||||
| Step function | 2049.07 | 0.00 | 1 | 1.00 |
| Sigmoid function | 2049.07 | 0.00 | 1 | 1.00 |
| Bilinear function | 2048.83 | 0.24 | 2 | 0.89 |