Projecting the time trend of thyroid cancers: Its impact on assessment of radiation-induced cancer risks

The incidence of thyroid cancer, which may be induced by ionizing radiation, has been rising in most Western countries for more than 20 years. In France, public worry about this increase and its possible connection with the fallout from Chernobyl led the government to ask for an evaluation of the health impact of this accident and an assessment of the feasibility of an epidemiological study. These requests raise two methodological questions: Which risk model should be used to relate exposure to risk? What is known about the spontaneous incidence rate of thyroid cancers? This article analyzes the impact of the time trend in the spontaneous incidence of thyroid cancers over the past 20 years in France when evaluating the risk of radiation-induced cancer. Age-period-cohort models were used to model the trend of spontaneous incidence from 1978 through 1997 and then to apply two scenarios for projections up to 2007: one with a constant incidence, the other using the trend observed over the past 20 years. Then the risk was assessed for a hypothetical population of 30,000 children aged 0 to 15 y, exposed to a hypothetical 0.1 Gy thyroid dose. The analysis shows that consideration of the trend instead of a constant spontaneous incidence can yield substantial differences in the risk estimates for thyroid cancer.