Chernobyl: Health Statistics for Adults, and a Counterfactual

By: Aaron Datesman

Many of the source documents referenced in Yablokov et al. (the NYAS Chernobyl report) may be found by poking around on the internet, which in this case is a highly worthwhile exercise. For instance, it's rather easy to locate a source for the chart in this post describing the juvenile morbidity in Gomel. The report is a product of the German Society for Radiation Protection (GFS) and the German affiliate of International Physicians for the Prevention of Nuclear War (IPPNW).

Having examined juvenile morbidity in one highly-contaminated region of Belarus, I agree that it's important to ask as well what the health outcomes have been for adults living in contaminated areas. The information below comes from page 62 of the GFS/IPPNW report.

The ultimate source for these statistics is Dr. Angelina Nyagu of Ukraine, whose biography indicates singular experience in the field of health effects due to the Chernobyl disaster. Dr. Nyagu's presentation at the April 2011 IPPNW Congress is available at this link, in case you are interested to learn more.

While these health statistics are quite frightening (the statistics for children of exposed parents especially), it's quite correct to note that this information includes no counterfactual. This is a substantial concern because, without health statistics representing an unexposed control population, it's impossible to disentangle negative health effects due to social and political instability in the Soviet Union circa 1990 from the negative health effects due to the Chernobyl disaster.

This information does exist, although I do agree that it doesn't get very prominent presentation in Yablokov et al., or in other studies released by anti-nuclear groups and cited by activists. It is possible to view this state of affairs as an attempt to create a non-existent causal relationship using an avalanche of data showing correlation without establishing cause. This is a reasonable critique up to a point, in my opinion, but in the case of the Chernobyl disaster it generally fails to pass the test of common sense.

There is a more basic explanation, however: robust, large-scale epidemiological studies are hellishly expensive. With the Soviet Union in a state of sustained upheaval and collapse in the years after the Chernobyl meltdown, who would have paid for suitable control studies? Furthermore, given the incentives to disguise the degree of harm which occurred (whatever it might have been), why would they have done so?

Nevertheless, counterfactual evaluations do exist. One such study is summarized in the graph below, taken from Yablokov et al. I cannot find the original reference (which is a Russian-language document, commissioned by Greenpeace) on the web, but the author appears to be Veniamin Khudolei of the Center for Independent Environmental Assessment of the Russian Academy of Sciences. (This article discusses the 2006 Greenpeace report which utilized the data shown below and was responsible, in turn, for the NYAS Chernobyl publication.)

Figure 7.24 displays the relative rates of mortality (per 1000 - that's 100 per 100,000, which is the more usual unit) in two areas of Russia contaminated by fallout from Chernobyl. In each year, the raw statistics have been adjusted (normalized) so that the value for Russia overall is zero, as indicated by the horizontal line. The graph therefore indicates that, prior to the Chernobyl disaster, the inhabitants of both Tula and Lypetsk a) enjoyed about equal mortality rates, and b) suffered less mortality than the population of Russia as a whole. Post-Chernobyl, mortality in both regions increased significantly with respect to the Russian baseline, with a much larger effect observed in Tula, which was more heavily contaminated.

A quantitative interpretation is also available. Starting in 1988, when the mortality rates in Tula and Lypetsk were identical, according to the figure an excess mortality emerges in Tula given by the length of the vertical blue line. By eyeball, on average the length of this line is about one unit, representing one death per year per 1000 population. Over the 25 years from 1998-2003, then, the excess mortality in Tula comes to 2.5% of the entire population of that contaminated region. (The actual value, more properly given by an integral representing the area between the curves, is 3.75% from 1990 to 2004).

According to Yablokov et al., that's 60,400 deaths in Tula alone.

Of course, I can't even locate Tula and Lypetsk on a map, so I have no idea whether the author of this study is performing the sort of sleight-of-hand I would call out if he were comparing health outcomes between Massachusetts and Mississippi. To my knowledge, however, those agencies and persons who dismiss the conclusions of Yablokov et al. do not do so from a foundation of deep knowledge regarding the demographics, geography, and health characteristics of the former Soviet Union. Given the obvious cultural, language, and political barriers to full communication regarding the effects of the Chernobyl disaster which exist, it is not apparent to me why the sensible conclusions of Russian health researchers and scientists can be blithely dismissed.

— Aaron Datesman