Risk Estimates for the
Health Effects of Alpha Radiation

by Duncan C. Thomas and K. G. McNeill

a report prepared for the
Atomic Energy Control Board (AECB)

published as
AECB Research Report INFO-0081
September 1982

with explanatory comments by Gordon Edwards

~ unabridged ~

This report provides risk estimates for various health effects of alpha radiation.

Human and animal data have been used to characterize the shapes of dose-response relations and the effects of various modifying factors, but quantitative risk estimates are based solely on human data:

  • for lung cancer, on miners in the Colorado plateau, Czechoslovakia, Sweden, Ontario and Newfoundland ;

  • for bone and head cancers, on radium dial painters and radium-injected patients.

Slopes of dose-response relations for lung cancer show a tendency to decrease with increasing dose.

    [ This means that linear extrapolation of health data from high-dose to low-dose situations tends to underestimate the actual number of radiation-caused lung cancers. ]

Our best estimate of curvilinearity is given by raising dose to the power 0.92 ▒ 0.07, but the improvement in fit beyond simple linearity is not significant.

On the other hand, the addition of a cell-killing term significantly improves the fit of the linear model.

In any event, linear extrapolation is unlikely to underestimate the excess risk at low doses by more than a factor of 1.5 .

    [ This means that linear extrapolation from high-dose data to low doses probably won't underestimate the number of radiation-caused lung cancers by more than fifty percent. ]

However, these inferences about curvilinearity are highly subject to error from the choice of reference populations, dosimetry, and latency.

Under the linear cell-killing model, our best estimate of excess relative risk is 2.28 ▒ 0.35 per 100 WLM (a doubling dose of 44 WLM).

    [ Thus , an accumulated dose of 44 working-level-months (WLM) will double a worker's risk of dying from lung cancer. Under the existing regulatory limit -- 4 WLM per year -- a worker could legally get this dose in 11 years. ]

Attributable risks in these five studies range from 3.4 to 17.8 per million person-years per WLM.

Risks from radon daughters appear to interact with age and smoking in a form intermediate between additive and multiplicative, though on balance, closer to multiplicative. We therefore favor the "relative risk model" for projecting lifetime risks, but have carried out life-table projections under a wide variety of assumptions.

Our best estimate of the effect of a 50-year occupational exposure to 4 WLM per year is 130 excess lung cancer deaths per 1000 persons (0.65 per 1000 person-WLM) with a range from 60 to 250 per 1000.

Similar calculations for lifetime exposure to an additional 0.01 WL beyond normal background produces an estimate of 10 excess lung cancers per 1000 persons.

Our risk estimates for radium are 2.4 bone sarcomas and 2.0 head carcinomas per hundred thousand PY-ÁCi   [ person-year-microcuries ] .

The lifetime risk from radium in drinking water at the Canadian maximum acceptable concentration is about 0.4 per 1000 persons, compared with a natural risk of 1.0 per 1000.   [ This represents a 40 percent increase in risk. ]

No major health effects of plutonium have yet been demonstrated in human populations, probably because of the small number of persons exposed to significant doses, though animal studies clearly show its carcinogenic potential.

Other effects of alpha emitters which have been reported include gastrointestinal, skin, and liver tumors, leukemia, liver cirrhosis, and chromosomal abnormalities, but these require further study before their risks can be adequately described.


~ excerpts ~

1. The lung is a major radiosensitive site, and for short-lived or insoluble inhaled alpha emitters   [ including radon gas ]   it is the primary concern.

2. Our best estimate for the excess relative risk   [ associated with radon ]   is 2.28 ▒ 0.35   [ times the normal incidence of lung cancer ]   per hundred WLM, which gives a doubling dose of 44 WLM.

3. Considering the differences in populations and methods. and comparing these results with those of other epidemiological studies (such as those on asbestos), there is a remarkable degree of agreement between the various studies   [ from Colorado. Czechoslovakia, Newfoundland, Ontario and Sweden ] . Only the Colorado plateau data [ where the total exposures were higher ]  stand out as giving very much lower risk estimates than the other studies of miners,

6. There is no justification for assuming that linearity would be a conservative basis for radiological protection for high-LET effects.   [ In other words, it cannot be assumed that the linear hypothesis -- which is used by the authorities -- will over-state the biological damage from alpha radiation; it may in fact under-state the damage. ]  

7. On the basis of a linear dose-response model with a constant relative risk of 2.28 per 100 WLM, the 4 WLM per year standard for occupational groups could increase the lung cancer risk of an individual working all his life at the maximum by a factor of from 2.4 to 4.2 (best estimate 3.8).

8. This is of course an estimate of the maximum risk obtained at 50 years of constant exposure to the maximum level of 4 WLM per year. It should be recognized that average exposures and hence average risks would be very much lower.   [ Thus, if average exposures are only one-tenth of the maximum, the 4 WLM per year standard would on average over the entire industry produce a 10 to 20 percent increase in the risk of lung cancer. ]  

10. On a similar basis, a 0.02 WL maximum   [ radon exposure standard ]   for homes could increase the lifetime lung cancer risk [ of people living their lives in such homes ]   by about 40 percent. However, this is the predicted increase for a lifetime of additional 0.02 WL exposure beyond normal background levels (which vary widely but might reasonably be as much as one third of 0.02 WL).

16. Protraction of dose   [ that is, delivering the same dose over a longer period of time ]   appears to increase risks of lung cancer (and other effects). For this reason, epidemiological studies of populations exposed for relatively short periods may underestimate the risks of life-long exposure.

~ excerpts ~

1. In view of   [ analytical uncertainties ]   these conclusions should be considered to be highly tentative. However, as the necessary data already exist for several mining populations, many of these limitations could be overcome relatively quickly by reanalyses.

We therefore recommend that linearity be treated only as a temporary ad hoc basis for interim standards, which should incorporate a "safety factor" to allow for the possibility of convexity     [  whereby actual risks are greater than those anticipated by the linear hypothesis ] ; this safety factor need not be more than two-fold. The conclusion should be reviewed after better analyses have been carried out on existing or improved data.

2. The Ontario and Newfoundland mining populations should continue to receive the best possible follow-up; in particular, efforts should be made to obtain smoking histories, to evaluate the contribution of radiation to the reported risks of gastrointestinal and skin cancer, and to use the best available statistical methods.

5. Consideration should be given to initiating a cohort study of family members of Ontario uranium miners during the 1950-1960 boom, and a case-controlled study of lung cancer in relation to radon daughter levels in homes. Of the two, the case-control study would probably be the more feasible starting point. We estimate that a sample size of about 500 cases and 500 controls would be sufficient to demonstrate a significant association with lung cancer risk. Should it prove feasible to identify and trace family members of Ontario uranium miners, similar sample size calculations could be easily done.   [ These recommendations have been totally ignored. ]

7. The criteria for awarding compensation for lung cancer in radiation workers should be re-examined in the light of the evidence contained herein, We suggest that on the average, 50 percent of the lung cancers with an accumulated exposure of about 44 WLM (the "doubling dose") would have been radiogenic in origin. The onus of establishing causation should not be placed on the individual concerned,

9. In our view, the present 4 WLM per year maximum does not provide sufficient protection to an individual who works at that level for an entire lifetime, though for an entire group of uranium workers the average risk resulting from the use of this individual maximum might be considered acceptable. We recommend that regulations setting permissible annual exposures be supplemented by additional explicit requirements that average exposures to all relevant (i.e. potentially exposed) workers in a company be very much lower.

It is also desirable that few workers ever accumulate total lifetime exposures approaching the maximum possible under present regulations (approximately 200 WLM). We therefore recommend that lower maximum individual doses be encouraged by education of the workers themselves, maintenance of low average doses, and non-discriminatory policies of job rotation.

Uranium miners facing higher risk
of lung cancer death, study says

By Barry Kliff
of The Montreal Gazette

The Atomic Energy Control Board (AECB) says there is no substantiation to a study it commissioned, which shows uranium miners face a greater risk of dying from lung cancer than the general public.

The $15,000 report commissioned by the board says Canada's 4,500 uranium miners are four limes more susceptible to lung cancer death.

It says the board should consider tougher controls on radioactive gases, especially radon emitted during mining, which is an apparent cancer cause.

But the board wants to ease controls on radon, said William Bush. manager of the board's radiation protection division.

Bush said workers are adequately protected. He said international studies have shown it is possible to raise the amount of radon gas a miner is exposed to while decreasing his over-all exposure to radioactive gases.

"There is no substantiation in the author's report for the conclusions they have drawn," Bush said. "I don't think anything would be gained from studying the matter any more."

One of the report's co-authors, Duncan Thomas, Associate Professor of Epidemiology at McGill University, said other international studies show uranium miners aren't adequately protected against radon gas.

To reach his conclusions, Thomas compared results of five studies involving 18,000 miners in Canada, Czechoslovakia, Sweden and the United States. It was not original research.

"If the AECB responds by raising the limit, I feel our work has been in vain." Thomas said.

Each year among the general public, Thomas said, five per cent of all deaths are due to lung cancer. But under today's standards for miners, Thomas said, the risk could reach 20 percent.

In two studies. involving a total of 5,500 miners in Newfoundland and the U.S., Thomas said, researchers expected to find 35 cases of death caused by lung cancer but found 224.

Bush said various AECB committees are studying Thomas' report now and a final report is expected early next year.

AECB disregards report saying
radon gas exposure standards
are inadequate

By Mitchell Beer
for Southam News

Page 12, The Ottawa Citizen,
Wednesday. October 20, 1982

The Atomic Energy Control Board has chosen to disregard a research report recently commissioned that concludes current standards for exposure to radon gas are inadequate.

The report found uranium miners exposed to the radioactive gas over a full working life could run a 20- to 30-percent risk of dying of lung cancer. compared to about five percent for the general population.

"The conclusion we arrive at . . . is that the standard does not provide adequate protection to an individual, and the only justification for such a high standard can be that the average exposure to the entire work force is very much lower, "said Duncan Thomas, Associate Professor of Epidemiology at McGill University and co-author of the report.

The report also found maximum standards for radon exposure in homes would give residents a 40-per-cent higher than average lung cancer risk.

William Bush, manager of AECB's Radiation Protection Division, said a board advisory committee has reviewed the available literature on radon gas. including the research report. "Their conclusion is that there is no basis for reducing the limit."

Thomas said the average risk a group of miners would face, might appear low, but "the maximum risk an individual can theoretically run with today's standards is not acceptable or negligible. It's certainly a measurable increase in risk."

Based on his findings, Thomas said the increased risk of lung cancer for miners and occupants of radon-exposed homes "are substantial enough that they ought not be dismissed, and serious consideration should be given to the possibility of lowering the standards." But Bush said AECB is now considering a 20-percent increase in allowable radon exposure, in line with changes in international standards.

Because the increase is part of an integration of standards for various types of radiation. the net effect is a tougher overall limit, Bush said. "In effect, you're reducing it, because people are never exposed to radon alone."

But Gordon Edwards. head of the Montreal-based Canadian Coalition for Nuclear Responsibility. said AECB staff are on record as saying radon exposure up to 100 times the current annual limit would be safe. Edwards said medical doctors have told him they know of no other carcinogen with such high permissible exposure levels.

"The AECB has for years been reassuring people that the standards are set so stringently that there's no significant risk," he said, "but according to the Thomas-McNeill findings, the risk is very significant."

[ Uranium Sub-Directory ] [ Plutonium Sub-Directory ]

[ Findings on Uranium Tailings ] [ COMPLETE DIRECTORY ]


Since March 27th 1996, there have been over
100,000 outside visits to the CCNR web site, plus

(counter reset June 3rd 1998 at midnight)