Sammendrag
This thesis is investigating if the risk of childhood leukemia and cancer in the central nervous system (CNS) increases in relation to high levels of indoor radon concentrations at home or low parental socioeconomic status. For assigning indoor radon exposure in dwellings lacking on-site measurements, we developed a new buffer model based on the distance to the nearest measured houses.
All 712,674 children born alive in the Oslo region in the period 1967-2009 identified from The Norwegian Population Register were included. Through geographical information systems the residence of every child was geo-coded and assigned a radon exposure value for each household. The model was based on 41,515 indoor radon measurements. We had on-site indoor radon measurements for only for 6% of the residences in the study region and therefore, we used a buffer model with different radius sizes to estimate radon exposure to the rest of the cohort. The model was validated against two datasets, the indoor radon measurements available in the study, and results from a regression model constructed with radiometric data and bedrock geology.
We took into account residential history for every child. Exposure was estimated for each year from birth to whatever came first: the age of 15 years, the date of diagnosis of leukaemia or cancer in the CNS, emigration, death or the end of the follow-up period in 2009. A total of 437 children got leukaemia and 427 got cancer in the CNS.
From The Medical Birth Register of Norway, we obtained several risk factors associated in earlier studies with the risk of childhood cancer, date of birth, sex, birth weight, gestational age and congenital malformations. For the mother, we had year of birth, parity, and complications during pregnancy, including haemorrhage and hypertensive conditions. Parents’ highest educational level and annual family income data were obtained from Statistics Norway.
Family income was categorized into quartiles and grouped into three categories, below the official poverty line (OECD-50 limit, medium income, and high income. We included information on marital status and others birth characteristics. Parents’ educational level was categorized in three subgroups, less than 10 years, 10-12 years and 13 or more years.
We used radon mean concentrations to analyze the radon exposure for each child at different points in times, during the first year of life, between birth and four years of age and in the whole study period (0-15 years).
Results
We found a reasonable good agreement between radon values from buffers compared to radon values from a regression model constructed with radiometric data. The level of agreement varied between 0.54 and 0.67. Comparing estimates from the buffer method with indoor radon values from measured houses had a level of agreement between 0.63 and 0.68. (Paper I).
No association was observed between radon exposures at home and the risk for leukemia (HR=1.00, 95% CI=0.87–1.15) or cancer in the CNS (HR=1.13, 95% CI=1.00–1.28) (Paper II).
An elevated risk of lymphoid leukemia among children in the lowest income category during the first two years of life was observed (OR=1.72, 95% CI=1.11-2.64). For cancer in the CNS we found no increased risk related to family income (Paper III).
Conclusions
The buffer method of estimating indoor radon exposure is suitable for use in epidemiological studies.
No association between radon exposure at home and the risk for leukemia or cancer in the CNS among children younger than 15 years of age living in the Oslo area was observed.
Being born into a household of low family income the first two years of life might be a risk factor for lymphoid leukemia.