Adjusteda incidence rate ratios for primary brain tumour in association with a 100 Bq/m3 increase in domestic radonb within strata of NOx at the residential address.

aWe adjusted the analyses for age (underlying time scale), sex, employment in the chemical industry for at least one year and consumption of fruit and vegetables.bRadon exposure was entered as a continuous variable in all models as the time-weighted average concentration at residences from 1. January 1971 until censoring.with inclusion of a 10 year latency period.cTest of the null hypothesis that the linear trends are identical, for Wald test for interaction.dTime-weighted average concentration for NOx.