Linear dose*-response modelling results by preferred lag and exposure window
| Outcome | No. cases | Adjusted for | Best fit exposure lag ERR per 100 mGy (95% CI) | Best fit exposure window ERR per 100 mGy (95% CI) | ||
|---|---|---|---|---|---|---|
| Lag (years) | Lag model | Window (years) | Window model | |||
| Non-CLL† | 264 | Sex, race, hire year | 2 | 0.09 (−0.17 to 0.65) | 6 to 14 | 1.9 (<0 to 8.0) |
| AML | 150 | Birth cohort | 1.1 | 0.22 (−0.19 to 1.2) | 6 to 14 | 7.0 (0.079 to 32.0) |
| CML | 52 | Sex, race, hire year | 6.3 | 0.29 (<0 to 2.6) | 6.3 to 14 | 3.8 (−1.2 to 26.0) |
| CLL | 74 | Sex, race, birth cohort | 24 | −0.032 (−0.16 to 0.74) | 11 to 14 | −2.8 (NC to NC) |
*‘Dose’ was defined as absorbed dose to red bone marrow from low linear energy transfer radiation exposure.
†Excluding leukaemias of indeterminate subtype.
AML, acute myeloid leukaemia; CML, chronic myeloid leukaemia; CLL, chronic lymphocytic leukaemia; ERR, excess relative risk; NC, not calculable.