WHO reports low health risk from Fukushima

17 May 2013

Two years on from the accident at Fukushima Daiichi in Japan, the World Health Organization (WHO) has published the first preliminary analysis of the global health effects due to radiation exposure from the event.

Its report concludes that "no observable increases" in cancer rates are anticipated for the general population as a result of the Fukushima Daiichi nuclear accident of March 2011. However, it found that the risk of developing certain cancers has marginally increased for some residents of Fukushima prefecture, and calls for long-term continued monitoring and health screening for them.

The study, titled 'Health Risk Assessment from the nuclear accident after the 2011 Great East Japan Earthquake and Tsunami based on preliminary dose estimation,' was released on 28 February. The study, which began soon after the accident, was intended to "support the identification of needs and priorities for public health actions."

The assessment was conducted by about two-dozen "independent international experts" in radiation risk modeling, epidemiology, dosimetry, radiation effects and public health, as well as representatives from the WHO. The United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR), which is conducting its own assessment of exposure levels and effects of the Fukushima accident on humans and the environment (expected late 2013), also participated in two expert group meetings as an observer to ensure compatible approaches and data sources for the two UN assessments.

The WHO study estimated the lifetime risk for all solid cancers combined, and also for three types of individual cancer (leukaemia, thyroid cancer and female breast cancer), which are most closely associated with radiation exposure. Risks were estimated for both sexes and for infants, children and adults. The age groups were chosen to allow characterization of risks for "younger and more sensitive populations," according to the report.

Heath risks for the Tokyo Electric Power Company (TEPCO) male emergency workers were also estimated for three different ages (20 years, 40 years, and 60 years). In addition, the study 'qualitatively assessed' non-radioactive health effects for those workers, such as thyroid diseases, cardiovascular diseases and lens opacities (or cataracts).

Geographically speaking, the study covered four distinct areas. The regions, defined based on the estimated doses received in the first year after the accident, include the two most affected regions (locations 1&2) of Fukushima prefecture (but outside the 20km evacuation zone), Japan, and the rest of the world. Actual risks, however, were only calculated for regions of Fukushima prefecture as an expert group deemed that risks for the rest of Japan and the world "would be much lower than the temporal and spatial fluctuations of the baseline cancer incidence."

The risk assessment process comprised four main steps: hazard identification, hazard characterization (dose-response assessment, exposure assessment and finally risk characterization. Each of these stages is explained in detail by a chapter in the 172-page report.

Hazard identification

The hazard identification process involved identification of the source term or the amount and type of radionuclides released during the accident, as well as identification of potential health hazards due to ionizing radiation.

Data on the amount of radionuclides released during the Fukushima accident and deposited around Japan was based on the WHO preliminary dose estimate, among others. The dominant contributors to the exposure were identified as iodine-131, in the first weeks after the accident, and caesium-134 and caesium-137 later on.

The potential hazard from the radionuclides was determined based on previous experimental and epidemiological studies, with most population-based cancer risk estimates coming from the Japanese atomic bomb survivor life span study (LSS). That study, which looked at cancer mortality and incidence in 120,000 individuals, found increased radiation-related risks for three of the four types of leukaemia, and for a large number of solid cancer sites.

The health risk assessment group chose to focus on leukaemia, thyroid cancer and breast cancer "because the influence of early age-at-exposure is particularly relevant." Leukaemia has often been the first malignancy to show up after radiation exposure, around 2-5 years later.

The Fukushima health risk assessment also considered the risk of "all solid cancers." The LSS study showed an increased risk of solid cancers, beginning "several years" after the bombings. Risks are generally thought to be higher among women, and among younger exposure victims. Non-cancer radiation effects were also assessed, including the risks of thyroid disorders, radiation induced lens opacities, (which can lead to vision impairing cataracts), circulatory diseases, infertility, and teratogenic (hereditary) effects.

Dose-response relationship

This second step of the risk assessment process examined the relationship between exposure to a particular agent and any resulting adverse health effects in humans.

The HRA expert group decided that the Linear No Threshold (LNT) model provided the "most reasonable" description of the relation between low-dose exposure to ionizing radiation and the incidence of cancers. The LNT rests on the assumption that biological damage, which, if repaired incorrectly, could lead to cancer, is directly proportional to dose throughout a relevant range of doses and dose rates.

"Although some dissenting views on the LNT have been expressed, it is thought to be a prudent basis for risk assessment," the report stated, adding that it has been chosen in an attempt not to under-estimate the risks.

Based on cancer incidence rates from a general population, the lifetime baseline risk (LBR), that is, the baseline probability of having a specified cancer up to the age of 89 was calculated. The LBR calculation includes two terms. The first term is the baseline cancer incidence rate in the at-risk (but unexposed) population calculated from age-at-exposure. The second term is the general survival curve population, since, to be blunt, dead people don't contract cancer.

The HRA expert group picked lifetime-attributable risk as the primary risk measure for the study. The lifetime attributable risk (LAR) specifies the probability of a premature incidence of a cancer attributable to radiation exposure in a representative member of the population.

As for LBR, mortality data was taken from the Japanese statistics website. Cancer incidence data were taken from the 2004 Japan Cancer Surveillance Research Group compilation of 31 population-based cancer registries in Japan. Unfortunately, at the time of the study no local cancer incidence rates were available for Fukushima prefecture, however cancer incidence in two neighbouring prefectures (Miyagi and Yamagata) and elsewhere in Japan was deemed by the expert group as "likely to be comparable."

Exposure assessment

The exposure assessment (step 3) looked at how much of the particular radionuclides the groups were exposed to, the exposure pathways and how the exposure occurred.

General population

Doses were considered for the general population (infant, child, adult) in four geographical areas. Effective doses in the first year ranged from 12-25 mSv for the most affected parts of Fukushima prefecture, to less that 1mSv in the rest of the world (see Table 1 and map).

Table 1: First year organ doses in the general population, mSv
Location (see map)ColonBreastBone marrowThyroidColonBreastBone marrowThyroidColonBreastBone marrowThyroid
Location 1a2223216325252595262726122
Location 2a222321342525255226272673
Location 3a121312171414142815151548
Location 4554165552555543
Locations 5-9/(5-10 for thyroid)555145552255539
Locations 10-14/(11-14 for thyroid)444114441855435
Rest of Fukushima prefecture33383331533331
Neighbouring prefectures111411151119
Rest of Japan111111111111
Rest of world<0.01<0.01<0.01<0.01<0.01<0.01<0.01<0.01<0.01<0.01<0.01<0.01

*a: Organ dose from March-June 2011 Note: Based on Tables 5&6 (pp42-43) of the WHO report.  


The estimated doses were provided in a WHO report released in May 2012. Doses within a 20-km radius around Fukushima Daiichi NPP were not assessed in the preliminary dose estimation, because people in that area were "rapidly evacuated" and such an assessment would have required "more precise data" than was available, the report said (but see box, criticism).

The HRA (health risk assessment) expert group decided to use organ doses, rather than effective doses to estimate heath risks. Organ doses were calculated for red bone marrow, thyroid, breast, and colon, as input data for the cancer risk models for leukaemia, thyroid cancer, breast cancer and all solid cancers, respectively (see tables). Estimation of doses beyond the first year was not carried out as this would have "resulted in a high degree of uncertainty," according to the report. Instead, for the purposes of calculating lifetime risk, the expert group agreed that the dose over the lifetime should be approximated to be twice as much as the first-year dose. This assumption, taking into account differences between the two accidents, was based on the Chernobyl experience, where the ratio of long-term to one-year dose is estimated to be 3, based on data from up to 20 years after the accident. (The study points out that publicly-available data shows that similar proportions of Cs-134 and Cs-137 were emitted by Fukushima. In contrast, more than twice as much Cs-137 as Cs-134 was emitted by Chernobyl. As a result, the fraction of lifetime dose emitted by Fukushima will be lower than Chernobyl.)


In addition to the general population, the heath risk assessment also included an exposure assessment for some 23,000 emergency and mitigation workers (5639 TEPCO employees and 17,533 contractors) that may have been exposed to radiation during response to the accident. Risk modelling calculations were only performed for male workers, which composed the "vast majority" of the emergency workforce.

An assessment was not carried out for other categories of emergency workers (for example, police and firefighters) as radiation dose information was not available within the timeframe of the study.

The HRA expert group decided that the health risk assessment for workers should be based on "a few plausible exposure scenarios," rather than on individual doses received, again over a one-year period. Results are presented in Table 2.

Table 2: Estimated organ doses for the four scenarios assumed for NPP workers (mSv)
ScenarioBone marrowColonThyroid 
155569% of workers (ca. 16,000)
2242414030% of workers (ca. 7,000)
3200200200Less than 1% of workers (ca. 200)
410010011,800Upper bound (a few workers)

*Note: Based on Table 10 (p.50) of the WHO report.

Risk characterization

In addition to lifetime attributable risk (LAR), which is associated with very large uncertainties as it integrates the expression of radiation induced risk over the whole lifespan (up to 90 years) the study also estimated the risk over the 15-year period after radiation exposure (not shown here).

General population

The study found that, for the general population, based on doses estimated to be substantially below threshold levels, deterministic effects (that is, tissue reactions) are not expected.

In summary the report concluded that the increases in the incidence of human disease attributable to the additional radiation exposure from the Fukushima Daiichi NPP accident are "likely to remain below detectable levels."

It also said that radiation doses from the damaged nuclear power plant are not expected to cause an increase in the incidence of miscarriages, stillbirths and other physical and mental conditions that can affect babies born after a nuclear accident.

For locations outside of Fukushima prefecture and around the world, the radiation-related cancer risks were "estimated to be much lower than the usual fluctuation in the baseline cancer risks."

The predicted magnitude of cancer risks was assessed for leukaemia, thyroid cancer, female breast cancer and all solid cancers combined. The risks were calculated over a lifetime and over the 15 years following the accident. The lifetime attributable risks (LAR) were quantitatively estimated only in the most affected parts of Fukushima prefecture. Some of the results are shown in Figures 2-5. Worker data in these figures has been complied by NEI from report data.

The results show the largest additional cancer risks among those exposed in infancy (leukaemia in males and solid cancers in females). The estimated increased risk for thyroid cancer for the worst-affected areas was up to 70% in baby girls. On that point, the report said, "Due to the low baseline rates of thyroid cancer, even a large relative increase represents a small absolute increase in risks." For example, the baseline lifetime risk of thyroid cancer for females is just 0.75% and the additional lifetime risk estimated in this assessment for a female infant exposed in the most affected location is less than 0.50%.

However the report did state that the high relative risk of childhood thyroid cancer "becomes more evident when risks are calculated over the first 15 years after the accident," and thus concluded that monitoring children's health is warranted.

Emergency workers

The Fukushima accident has "not resulted in any acute radiation effects among workers." The study said that none of the seven deaths reported around the time of the accident were attributed to radiation exposure. It did note, however, that transient thyroid dysfunction was reported in three workers as a result of repeated self-administration of stable potassium iodide for thyroid blocking against radioactive iodine.

The potential health consequences of the exposure to radiation of the emergency workers were assessed for four different scenarios (Table 2). The assessment concluded that "no deterministic effects of radiation are expected in the workers, apart from possible thyroid disorders in those few workers who inhaled significant quantities of radioactive iodine."

In terms of cancer risks, the study, which considered workers aged 20, 40 and 60 years, found that risks were "consistently lower" for workers exposed at an older age (over half of the workers were aged between 40-59 as of January 2012). For around two-thirds of the emergency workers (Scenario 1), all calculated risks are of similar magnitude as the normal fluctuations in the baseline cancer risks. For about one third of the workers (Scenario 2), the relative increase over background for thyroid cancer is estimated to be up to 20% for the youngest workers. For less than 1% of workers (Scenario 3), the relative increase over background for leukaemia and thyroid cancer is as high as 28% in the youngest workers. For those few emergency workers who received very high doses to the thyroid (Scenario 4), a "notable risk" of thyroid cancer is estimated, especially for young workers.

Health considerations

The report said that the accident "highlights the need for continuing and improving low-dose and low-dose-rate radiation research."

It also called for "surveillance of health and monitoring of disease occurrence." It said that this is required for "empirical assessment" of the health consequences of the accident and to quantify the health outcomes resulting from it. It will also "permit the identification of additional needs for the delivery of health care."

The report highlighted mental health risks as "a challenge" to the medical community, although that was beyond the scope of the current risk assessment.

Potential sources of uncertainty

Dose estimates

Conservative assumptions were made about protective measures and food consumption that were likely to "overestimate" radiation dose.

Dose estimates were modelled using environmental/food data, rather than being based on actual human measurements.

Organ dose was used in the cancer risk models. However, the ratio between absorbed dose and effective dose is dependent on the radionuclide involved, leading to some intrinsic uncertainty in the calculations.


Cancer data from Fukushima were assumed to be comparable to those from other parts of Japan.

The expert group estimated the long-term dose from Fukushima would be twice as much as the first-year dose, based on experience of the Chernobyl accident (and taking into account differences).

Risk models

The proposed model structure and parameters are based on assumptions derived from uncertain values.


Based on "Health risk assessment from the nuclear accident after the 2011 Great East Japan earthquake and tsunami." (http://tinyurl.com/whohra)

Figure 1
Figure 3
Map of Cs-137 deposition in Fukushima prefecture
Figure 4
Figure 2

Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.