Incidence of thyroid cancer in Scandinavia following fallout from atomic bomb testing: an analysis of birth cohorts

Objectives: The occurrence relation between radioactive fallout from nuclea r testing at Novaja Semlja in north-west Russia and the incidence of thyroi d cancer in Norway and Sweden was studied following a birth cohort approach . Methods: Birth cohorts with presumably different levels of exposure were id entified according to calendar year of atomic tests and previous Norwegian estimates of the population dose (born 1947-1950 received low exposure in l ate childhood, born 1951-1962 received the highest exposure in early childh ood, born 1963-1970 were not exposed). For each one-year birth cohort the i ncidence rates were calculated, with denominators based on exact population figures for each year of follow-up. Results: In a stratified analysis, the relative risk for the highest expose d cohorts born 1951-1962, compared to those not exposed born 1963-1970, was found to decrease with increasing age from a borderline significant relati ve risk (RR) of 1.7 (95 percent confidence interval, 95% CI: 1.0-3.0) for c hildren in the age-group 7-14 years to no excess risk among those 20-24 yea rs of age (RR: 0.9; 95% CI: 0.7-1.2). The mean age at diagnosis of thyroid cancer in the age-group 7-14 years was lowest in the birth cohorts with the highest exposure. The Poisson regression analysis showed essentially the s ame results, with an improved fit when adding an interaction term between a ge and birth-cohort to a basic model with age, gender, birth-cohort and cou ntry. Conclusion: These results are compatible with an increased risk of thyroid cancer during childhood and adolescence for subjects exposed to radioactive fallout early in life. Alternative explanations for the pattern of inciden ce are discussed.