A cohort grows older
Published 12 October 2026
The 2018 thyroid-cancer assessment followed work begun while the arch was being completed. In mid-2016 UNSCEAR asked its secretariat to evaluate thyroid-cancer data from the affected regions, and the resulting white paper was published in 2018. It drew on official incidence data from Belarus, Russia and Ukraine and on scientific literature through December 2016. The total number of thyroid-cancer cases registered from 1991 to 2015 in people who had been under eighteen in 1986, across the whole of Belarus and Ukraine and the four most contaminated Russian regions, exceeded 19,000. That was almost three times the 6,848 cases the committee had recorded for the same cohort through 2005. The registered total had grown as the exposed cohort aged and as diagnosis continued. C003, unscear.org, UNSCEAR 2008 Report - Annex D (corr) C063, unscear.org, UNSCEAR Chernobyl white paper, 2017 C069, unis.unvienna.org, UN Scientific Committee launches a new evaluation of thyroid cancer data in regions...
The white paper was explicit that the rising count had several causes. It listed the natural increase in spontaneous thyroid-cancer incidence as the cohort aged, radiation exposure, greater awareness of the risk after the accident, and improved diagnostic methods. Screening had a large effect on measured incidence: in one Ukrainian-American cohort the crude rate was many times the regional rate for the same birth cohort and period, largely because intensive examination finds cancers that ordinary care misses. The committee found no evidence that screening significantly influenced excess relative risk per unit thyroid dose; confidence bands for the UkrAm and Japanese atomic-bomb survivor cohorts overlapped despite their different screening conditions. A bigger case count, in other words, did not mean the radiation effect had tripled; detection itself was part of the increase. C063, unscear.org, UNSCEAR Chernobyl white paper, 2017
Against that background the committee offered an estimate, not a count, of attribution. For non-evacuated residents of Belarus, Ukraine and the four contaminated Russian regions who had been children or adolescents in 1986, it assessed the fraction of thyroid-cancer incidence attributable to radiation exposure for 2001 to 2008 as of the order of one quarter. The estimate rested on assumptions: a linear dose-response across the relevant range and the applicability of dose-risk figures from cohort studies to the wider population. The statistical uncertainty alone put the credible range from about 0.07 to 0.5, and the true uncertainty was wider once dose estimation and the linearity assumption were included. The fraction had probably been higher in the first years after the latency period and would decline slowly afterwards. C063, unscear.org, UNSCEAR Chernobyl white paper, 2017
The United Nations public summary of the white paper translated that fraction into a headline: about 20,000 registered cases in the cohort, of which the committee estimated one in four was attributable to radiation exposure. That formulation is a shorthand for an estimated fraction applied to a registered total within defined geographic and age boundaries. It is not a registry of five thousand individually diagnosed radiation cancers, and the committee itself stressed that no biomarker could yet identify which individual cancer was radiation-related. The careful version of the finding is a population-level estimate with a wide uncertainty range, derived from dose-response evidence and assumptions the paper states. C063, unscear.org, UNSCEAR Chernobyl white paper, 2017 C069, unis.unvienna.org, UN Scientific Committee launches a new evaluation of thyroid cancer data in regions...
The white paper's closing list of open questions shows how much remained unsettled three decades on. The committee still needed to quantify the risk from thyroid doses below 500 milligrays, to establish how long the radiation-induced risk persists and how it declines, to understand confounding factors such as iodine deficiency, and to find a biomarker that could distinguish a radiation-induced thyroid cancer from a spontaneous one. Each of those questions bears on numbers the public argument treats as settled. The attributable fraction of one quarter is a central estimate inside a wide range, and the case total it applies to is still rising as the cohort is followed. The honest statement of the evidence in 2018 was a strong association, a quantified but uncertain attribution, and a research agenda, in that order. C063, unscear.org, UNSCEAR Chernobyl white paper, 2017