Nuclear fallout and lung cancer

"On the basis of data available from two representative samples of lung cancer deaths in the United States as well as national mortality statistics and other epidemiologic studies, the lung cancer mortality rate has risen substantially between 1914 and 1968 among persons who never smoked cigarettes. For white males the relative increase for ages 35--84 years has been about 15-fold; the relative increase for ages 65--84 years has been about 30-fold. For white females the relative increase for ages 35--84 years has been about seven fold. Most of the relative increase occurred before 1935 and was probably due to changes in diagnostic criteria. However, increases have continued up to the present for male nonsmokers, who now apparently have an annual age-adjusted lung cancer death rate of about 25 per 100,000 persons between the ages 35--84 years. The rising lung cancer rate among nonsmokers indicates that factors in addition to personal cigarette smoking have had a significant effect on the mortality rate from this disease. In spite of the limited quality of these data, they suggest that a more complete understanding of lung cancer etiology is needed."    Rising lung cancer mortality among nonsmokers. Enstrom JE.
The theory that the post world war rise in lung cancer was caused by rising cigarette consumption can not possibly explain the rise of lung cancer deaths in persons who have never smoked. However the period of nuclear fallout that occurred between 1945 and 1985 (UNSCEAR ANNEX C) can explain this as people would be exposed to inhaled radionuclides produced in atomic weapons testing regardless of their smoking status.

To test whether there is any correlation between the period of nuclear fallout and lung cancer rates , I have taken cancer data made available by the IARC and created charts to show the deaths that occurred during the period in question. Deaths that occurred during the nuclear fallout period are in red and deaths that occurred outside of this period are in black. If the global lung cancer epidemic was caused by nuclear weapons testing then we would expect to see a rise in lung cancers between 1945 and 1985 followed by a fall after 1985.

                                                                 The world age standardised rate (W)

Male and female the world age standardised rate (W)

Male and female the world crude rate

United States of America. Population 324,118,787
227. "The radioactive fall-out is cleared, sooner or later, from the air by deposition. Rain contains the bulk of deposited activity and continuing measurements have been made since 1951 of the radioactivity of rain water collected from specially treated roofs. Any radioactive dust deposited on the roofs in spells of dry weather is washed off and included with the next sample of rainwater. From these measurements the amount of radioactivity deposited per square mile can be determined for each explosion." - Medical Research Council report to UK government 1956

Brazil. Population 209,567,920

Russian Federation. Population 143,439,832

It is worth noting that in the USSR/Former Soviet Union smoking rates remained stable over the decades in question and yet the lung cancer signatures show a fall after atomic weapons testing fallout ends. If it were true that these epidemics were caused by smoking then this would be impossible.

Mexico. Population 128,632,004

Japan. Population 126,323,715

Philippines. Population 102,250,133

Egypt. Population 93,383,574

Germany. Population 80,682,351

Thailand. Population 68,146,609

United Kingdom. Population 65,111,143

"The only conditions in which an increased incidence of lung tumours
has been observed in association with radiation are those, in which there is
an increased risk of inhaling radon and the other daughter-products of
radium. In theory, however, the inhalation of radioactive material in particulate form, either as a result of fall-out from nuclear weapon explosions or in the vicinity of nuclear reactors, could lead to the accumulation of a high radiation dose within the lungs. Such particles would not be uniformly distributed within the lungs but would tend to aggregate on discrete small areas of the bronchi, which would thus be subjected to a high radiation dose, with the result that in the long run lung cancers might be produced in some people. In this country appropriate measures are always taken to eliminate the hazard in the vicinity of nuclear reactors, and it would be extremely unlikely to occur as a result of fall-out except in conditions of actual warfare. There is no evidence that external irradiation by X- or gamma rays can cause lung tumours in man" - Sir Harold Himsworth (Medical Research Council) report to HMG (United Kindom) June 1956

This 1956 report by Sir Harold Himsworth makes no attempt to estimate how many people could be killed by lung cancer as a result of atomic weapons testing when the report is littered with estimates for less politically sensitive risks. It also makes the bizarre claim that this risk would only materialise from weapons used in war because presumably alpha emitting radionuclides only cause lung cancer in times of war. In 1955 the UK had a stockpile of 14 nuclear weapons and there have been a total of 520 atmospheric nuclear tests in the world so far.

Sir Harold Himsworth was head of the Medical Research Council 1949 - 1968.

A year later in June 1957 the UK Government accepted
"The link between smoking and lung cancer is one of 'direct cause and effect', a special report by the Medical Research Council"
"The government has indicated that an educational campaign to raise awareness on the dangers of smoking will be launched via local health authorities." - BBC June 1957
 This was just 30 days after the UK tested a 3 MT hydrogen bomb in operation Grapple.

So the UK Government decided not to educate the public about the risk of lung cancer (and therefore deaths) from atomic weapons testing but instead decided to 'educate' the public about smoking and lung cancer. Bearing in mind that rolling papers for smoking tobacco were invented in 1532 , it would seem this warning was a little late for a lung cancer epidemic that was about to occur some 400 years later. (atomic weapons testing had started over a decade earlier in 1945 with the trinity test.)

France. Population 64,668,129

Italy. Population 59,801,004

England and Wales. Population 56,640,000

South Africa. Population 54,978,907

Republic of Korea. Population 50,503,933

Colombia. Population 48,654,392

Spain. Population 46,064,604

Ukraine. Population 44,624,373

Argentina. Population 43,847,277

Poland. Population 38,593,161

Canada. Population 36,286,378

Peru. Population 31,774,225

Venezuela. Population 31,518,855

Uzbekistan. Population 30,751,602

Australia. Population 24,309,330

Romania. Population 19,372,734

Chile. Population 18,131,850

Kazakhstan. Population 17,855,384

The Netherlands. Population 16,979,729

Guatemala. Population 16,672,956

Ecuador. Population 16,385,450

Czechoslovakia. Population 15.6 million (1993)

Cuba. Population 11,392,889

Belgium. Population 11,371,928

Greece. Population 10,919,459

Dominican Republic. Population 10,648,613

Czech Republic. Population 10,548,058

Portugal. Population 10,304,434

Azerbaijan. 9,868,447

Sweden. Population 9,851,852

It is worth noting that in Sweden smoking became popular decades after the USA and yet both lung cancer epidemics start when atomic testing starts and not after any theoretical time lag associated with the rise of cigarette use.

Hungary. Population 9,821,318

Belarus. Population 9,481,521

Serbia. Population 8,812,705

Tajikistan. Population

Austria. Population 8,569,633

Switzerland. Population 8,379,477

Israel. Population 8,192,463

China, Hong Kong. Population 7,346,248

Bulgaria. Population 7,097,796

Paraguay. Population 6,725,430

Nicaragua. Population 6,150,035

El Salvador. Population 6,146,419

Kyrgystan. Population 6,033,769

Singapore. Population 5,696,506

Denmark. Population 5,690,750

Finland. Population 5,523,904

Turkmenistan. Population 5,438,670

Slovakia. Population 5,429,418

Scotland. Population 5,295,000

Norway. Population 5,271,958

Norway is an interesting case because as can be seen on the chart below, manufactured cigarette consumption peaks in Norway in 2004 at 3 per person, some 78 years after the USA last had that level in 1926. And yet both countries have very similar male lung cancer signatures. It is a strange notion of cause and effect where the effect comes before the cause. This is another situation where the cigarette hypothesis not only has to make assumptions to be true but has to make an assumption that is impossible  to be true - specifically that cause can come after effect in Norway but not in the USA.  The correlation between atomic weapons testing fallout and male lung cancer in both countries remains the stronger hypothesis on this evidence.

Costa Rica. Population 4,857,218

Ireland. Population 4,713,993

New Zealand. Population 4,565,185

Croatia. Population 4,225,001

Republic of Moldova. Population 4,166,463

Kuwait. Population 4,007,146

Panama. Population 3,990,406

Georgia. Population 3,979,781

Uruguay. Population 3,444,071

Armenia. Population 3,026,048

Albainia. Population 2,903,700

Lithuania. Population 2,850,030

Macedonia. Population 2,107,000

Slovenia. Population 2,069,362

Latvia. Population 1,955,742

Northan Ireland. Population 1,811,000

Estonia. Population 1,309,104

Mauritius. Population 1,277,459

Cyprus. Population 1,176,598

Tinidad and Tobago. Population 1,364,973

Luxembourg. Population 576,243

Suriname. Population 547,610

Malta. Population 419,615

Belize. Population 366,942

Iceland. Population 331,778

Barbados. Population 285,006

China: selected rural areas. Population no data

China: urban areas. Population no data

China: selected areas. Population no data


  1. Some French bloke26 December 2015 at 17:41

    A few remarks on the use of ASRs (Age-standardised rates):
    Japan is a highly developed country, with presently the highest life spans on record, and the use of ASR (W) pits it against the world population, including many countries with far lower life expectancies such as India and dozens of sub-saharan African countries, which are generally not included in the IARC database, and where most of the increase in world population occurred during the time-scales represented in the graphs. Consequently, the ASR graphs for developed countries like Japan, and the general ASR graph for 89 countries, are bound to reflect the effect of this adjustment more than any other factor.
    From the IARC database, here are the trends in lung cancer mortality in Japan from 1975 - a time when crude and standardised rates were similar - to 2013:

    Men: 19.66 (crude), 19.78 (ASR)
    Women: 7.19 (crude), 5.96 (ASR)
    Both sexes: 13.43 (crude), 12.87 (ASR)

    Men: 84.23 (crude), 27.19 (ASR)
    Women: 31.72 (crude), 7.80 (ASR)
    Both sexes: 57.98 (crude); 17.5 (ASR)

    China may be included, but if you check the relevant population pyramids, those 'selected areas' only total about 118 million people, less than 8.5% of the Chinese population, so the IARC figures for such a huge country are not very helpful. Moreover the criteria used to define 'rural' as opposed to 'urban' areas are not specified, and we can only speculate on how living conditions in rural China have changed over the last four decades.

    A final note on the averaging of the figures for both sexes: in addition to striking differences between national LC rates (evened out by the application of ASRs), and synchronical differences in male vs female rates, lung is the ONLY cancer site which shows not only divergent but sometimes *opposing* secular trends in female vs male rates, especially since the 1980s and the general decline in smoking prevalence, and I find that aspect of the problem mighty hard to reconcile with either the cigarette hypothesis or the fallout theory.

  2. I echo the reservations of Some French Bloke about sex averaging. It is one of my pet hates. Whilst I agree with him that the divergent and opposing trends in female vs male rates of lung cancer makes a nonsense of the cigarette hypothesis, I am less inclined to dismiss fallout just yet.

  3. The point about both sexes is well taken. I shall publish a chart showing male and female trends for the world when I get a chance.

  4. This comment has been removed by the author.

  5. I have added a chart showing male and female rates for the world. It is the second chart down from the top.

  6. Some French bloke30 December 2015 at 17:08

    Seems that the new chart illustrates the point in the last paragraph of my above comment rather well...
    Some additional observations on the power of age-standardised rates (ASRs) for propaganda purposes:
    The overall cancer death rate in Japan 2013 was 351.1/100,000, whittled down to a third (117.6/100,000) if ASR is applied. It even cuts their prostate cancer death rate for that year by over 74%: from 18.7/100,000 to 4.8/100,000! Quite handy if one is in the business of "proving" that the use of the prostate specific antigen test from the 1980s onwards has had a positive impact on mortality, even though the prostate cancer death rate for Japan in 1984 was at 3.95/100,000 (3.15 ASR). If the crude figures are too embarrassing to mention, just posit that the moderate increase would have been far greater "due to an ageing population"! The Mainstream Media routinely use those adjusted figures (sometimes not even specified in the small print) to make their point, in the assurance that only a minute fraction of the general public will bother to check and reevaluate the various 'findings' and advice presented in the article.
    These two graphs show how ASRs can serve not only to artificially lower mortality levels, but also to reverse a trend:

  7. Rather annoyingly blogspot won't embed images in posts which would show the difference between crude and ASR rates for Japan (which is very striking).

    I have added the crude rates for male and females for the world (third chart down) which also illustrates the point by comparing it with the ASR chart above it. It is worth noting that not all the data for all countries will be up to date so the trends for the last five years are so will probably not look like that in a few years time.