Health Risks of Radon and Other Internally Deposited Alpha-Emitters BEIR IV (1988) / Chapter Skim
Currently Skimming:
2 Radon
2 Radon
Pages 24-158
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 24... ...
As illustrated in Figure 2.1, radon decays with a half-life of 3.82 days into a series of solid, short-lived radioisotopes collectively referred to as radon daughters or progeny (Figure 2-1) (see Annex 2B)
|
From page 25... ...
Because of their wide distribution, radon daughters are a major source of exposure to radioactivity for the general public, as well as for special occupational groups. The estimated dose to the~bronchial epithelium from radon daughters far exceeds that to any other organ from natural background radiation.20 The recent recognition that some homes have high concentrations of radon has focused concern on the potential lung-cancer risk associated with environmental radon.
|
From page 26... ...
research, described in detail in the appendixes of this report, is briefly summarized below. DOSIMETRY By convention, the concentration of radon daughters is measured In working levels (WL)
|
From page 27... ...
and dose to target cells and tissues in the respiratory tract is extremely complex and depends on both biological and nonbiological factors.20 Because of differences in the circumstances of exposure, it cannot be assumed a priori that exposure to 1 W[M in a home and to ~ W[M in a mine will result in the same dose of alpha radiation to cells in the target tissues of the respiratory tract. Thus, an understanding of the dosimetry of radon daughters in the respiratory tract is essential for extrapolating risk estimates derived from epidem~ological studies of miners to the general population in indoor domestic environments.
|
From page 28... ...
The specific mixture of radon daughters also affects the dose to target cells, but to a smaller extent. The amount of radon daughters inhaled varies directly with the minute ventilation, i.e., the total volume of air inhaled In each minute.
|
From page 29... ...
. Exposures of animals to radon and its daughters have confirmed that exposure to radon daughters causes lung cancer (see Appendix HI)
|
From page 30... ...
THE COMMITTEE'S APPROACH TO ESTIMATION OF LUNGCANCER RISK Evaluation of the lung-cancer risk associated with radon daughters was the most challenging task faced by the committee. Numerous studies of underground miners exposed to radon daughters have shown an increased risk of lung cancer, in comparison with unexposed populations.
|
From page 31... ...
THE COMMITTEE'S ANALYSIS OF THE RISK OF LUNG CANCER ASSOCIATED WITH EXPOSURE TO RADON PROGENY The committee's risk estimates for radon-daughter exposures are based largely on its own reanalysis of the four principal data sets on
|
From page 32... ...
For the first three of the cohorts, we obtained data on individual miners; for the Colorado cohort, we were able to obtain only detailed summaries of the type described below. Some of the miners in the Colorado cohort were occupationally exposed to radon progeny prior to their employment in uranium mines.
|
From page 34... ...
Detailed discussion of our statistical models and methods and of their application to these cohort data is given in Annex 2A. Undoubtedly, many factors influence the occurrence of lung cancer in miners exposed to radon daughters.
|
From page 35... ...
other than cumulative exposure. The 5-yr lag period used in evaluating cumulative dose represents the assumption that increments in exposure have no substantial effect on the risk of lung-cancer mortality for at least 5 yr.
|
From page 36... ...
The excess relative risk, for a given cumulative exposure, decreased substantially with an increase in each of these variables. For prolonged exposures, as is the case for much of the cohorts' experience, a risk model that incorporates time since cessation of exposure was felt to overemphasize the date of the last recorded exposure, in contrast with the full exposure history.
|
From page 37... ...
Detailed support of the procedures and conclusions is provided in Annex 2A. In particular, the decline in excess relative risk with both age at risk evaluation and time since exposure, with these two factors acting largely independently of one another, was reasonably consistent over the cohorts.
|
From page 38... ...
Note that r(a) is the lung-cancer mortality rate from all causative agents, not just that due to radon exposure alone.
|
From page 39... ...
The most direct interpretation of the committee's mode! is as a summary of the total lung cancer experienced by the four cohorts of underground miners.
|
From page 40... ...
The confidence intervals represent only basic sampling variation, assuming, for example, no errors in expm sure assessment, and ascertainment of death from lung cancer. The confidence interval for all cohorts together (Table 2-2)
|
From page 41... ...
. These confidence intervals reflect only the uncertainty due to basic sampling variation and not that due to systematic errors in
|
From page 42... ...
The risk coefficients were derived from analyses of four data sets; random or systematic errors in the original data might have biased the risk coefficients in the recommended TSE model. Several statistical models were evaluated in the development of the TSE model.
|
From page 43... ...
The exposure variable that was entered into the mode} represented only occupational exposure to radon daughters.
|
From page 44... ...
Differential misclassification is of some concern. The diagnosis of lung cancer in uranium miners might be more vigorously pursued by clinicians who are aware of the association between lung cancer and underground mining.
|
From page 45... ...
If data on individual smoking habits among Al four miner cohorts were available, the comm~ttee's analyses could have been controlled for smoking as a cofactor in the risk of lung cancer. But such information is available only for the Colorado miners, and Appendix VIT presents the results of analyses for this cohort that consider combinations of smoking and levels of cumulative exposure.
|
From page 46... ...
Moreover, the estimate of the declining risk with time since exposure is based on the estimated effect of exposure, obtained by comparing miners at different levels of cumulative exposure. Smoking cessation would explain this effect, only if the pattern of cessation has been associated with both cumulative exposure level and time since exposure.
|
From page 47... ...
As was noted in the section, "The Comm~ttee's Analysis of the Risk of Lung Cancer Associated with Exposure to Radon Progeny" above, the multiplicative standard error due to sampling variation alone is about 30%0. It is of some value to consider what would be the joint effect on the comm~ttee's risk estimates of the six sources of uncertainty considered, if each had this degree of variation.
|
From page 48... ...
This section first reviews each of these sources of uncertainty and then presents the lung-cancer risk projections for exposures in mining and indoor environments. UNCERTAINTIES IN THE RISK PROJECTIONS Gender The committee's mode} is based only on data on males; with the exception of the small studies of indoor exposure, epidemiological data on females exposed to radon daughters are unavailable.
|
From page 49... ...
Thus, this analysis shows that, among the atom~c-bomb survivors, the absolute radiogenic excess is similar in males and females. Because Jung-cancer data on females exposed to radon daughters are not available, the committee cannot formally support its preference for a sex-specific relative-risk model.
|
From page 50... ...
Cigarette Smoking As discussed in Appendixes V and VIT, cigarette smoking causes widespread and extensive changes in the lungs and is the predominant cause of lung cancer. It was thus essential for the committee to provide risk projections for both smokers and nonsmokers.
|
From page 51... ...
If the comm~ttee's assumption of a multiplicative interaction is incorrect, the most likely alternative would be in the direction of a submultiplicative interaction. As discussed below, if a submultiplicative mode} is correct, use of a multiplicative mode} would overestimate the risk associated with exposure to radon progeny for smokers and, more substantially, underestimate the risk for nonsmokers.
|
From page 52... ...
Thus, the committee has assumed, in the risk projections described below, that exposure to 1 W[M in a home and exposure to 1 WI.M in a mine have equivalent potency in causing lung cancer. In situations where, as outlined in Annex 2B, detailed information on the domestic environment allows a more exact estimate of relative potency, it would be appropriate to scale the committee's risk estimates.
|
From page 53... ...
Note that L0 - Le is the average for a population of exposed persons. The number of years of life lost by those who actually die of lung cancer is usually much greater—tvoicalIv.
|
From page 54... ...
Because those risk factors affect current and future baseline lung-cancer rates, they also influence excess risks and risk projections associated with radon exposure. The committee simplified lifetime risk estimates by considering only cigarette use and by assuming a steady-state pattern of tobacco consumption.
|
From page 55... ...
is the sex-specific Jung-cancer mortality rate in Table 2-3. On the basis of this model, the lifetime risks of lung cancer and life expectancies for males are 0.0112 and 70.5 yr, respectively, for nonsmokers and 0.123 and 69.0 yr, respectively, for smokers.
|
From page 57... ...
of lung-cancer mortality for life exposure to radon progeny at constant rates of annual exposure. 69.7 yr, respectively, for males and 0.025 and 76.4 yr, respectively, for females.
|
From page 58... ...
COMPARISON OF LUNG-CANCER RISK PROJECTIONS FOR F EMALES As noted above, an alternative to a simple multiplicative model for lung cancer in females can be developed that is based on the experience of the Japanese atomic-bomb survivors (Equation 2-7~. Figure 2-6 compares lifetime risk estimates for females with the two approaches.
|
From page 59... ...
were correct. INTERVAL EXPosuRE TO RADON DAUGHTERS Exposure to radon daughters usually does not take place at a constant rate over a lifetime.
|
From page 60... ...
females for lifetime exposure to radon progeny. Comparison of models for interaction between sex and exposure (see text)
|
From page 61... ...
CONDITIONAL LIFETIME EFFECTS EXPOSURE FROM BIRTH AND FROM AGE KNOwN ALIvE Lifetime risk of lung cancer and expected years of life lost are related to a specified exposure profile for persons who are followed from birth. Those measures do not, however, provide information about lung-cancer risk after survival to some specified age, for example, in uranium miners who are free of disease at retirement or homeowners who are currently healthy but want to assess the consequences of lowering radon concentrations in their homes.
|
From page 64... ...
64 ·~ C4 ._ E~ o _ oo _ 0 cr ~ oo oo ~ o ~ o.
|
From page 65... ...
65 ~ ~ ~ t— ~ ~ — ~n oo _ r~ ~ O ~ ~ ~ ~ O ~ ~ ~ ~ O ~ oO O o ~ o~ oo ~ ~ ~ ~ ~ ~ ~ ~ o oo ~ _ ~ ~ ~ ~ o _ ~ ~ ~o o _ ~ _ o o o o o o ~ _ _ _ _ _ o ~ ~ ~ ~ _ ~ _ ~ ~ ~ ~ ~ ~ _ o o o o o o o o o o o o o o o o o o o o o o o o o o o o 0 cr~ o~ oo oO _ o o o o .
|
From page 69... ...
to ~ ~ ~ - ~ ~ ~ - (~} {~)
|
From page 71... ...
71 _ oo ~ O ~ ~ ~ 00 0O — ~ 00 0 — O ~ ~ ~D ~ ~ O ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 0 oo ~ ~ ~ r~ 0 o o o o o o o o o o o o o o - - - - o o o ~ ~ ~ .
|
From page 72... ...
72 C~ au U
|
From page 74... ...
0.10 69.67 71.51 73.28 77.11 86.25 0.20 69.61 71.45 73.22 77.07 86.25 0.50 69.45 71.29 73.05 76.97 86.24 1.00 69.19 71.02 72.77 76.80 86.22 5.00 67.24 69.01 70.71 75.52 86.04 10.00 65.14 66.85 68.49 74.12 85.82 20.00 61.82 63.43 64.99 71.82 85.38 Le —Leb (yr) 0.10 0.05 0.04 0.02 0.00 0.00 0.20 0.11 0.08 0.04 0.01 0.00 0.50 0.27 0.19 0.10 0.02 0.00 1.00 0.53 0.37 0.20 0.03 0.01 5.00 2.48 1.71 0.88 0.14 0.01 10.00 4.58 3.06 1.51 0.23 0.02 20.00 7.90 4.98 2.27 0.33 0.01 NOTE: Exposure rate assumed constant starting at birth.
|
From page 75... ...
0.10 76.42 77.83 78.62 81.06 86.87 0.20 76.40 77.80 78.59 81.04 86.87 0.50 76.32 77.72 78.52 81.01 86.87 1.00 76.20 77.60 78.39 80.95 86.88 5.00 75.22 76.60 77.39 80.46 86.88 10.00 74.05 75.40 76.21 79.88 86.88 20.00 71.88 73.20 74.01 78.76 86.86 Le —Leb (yr) 0.10 0.03 0.01 0.01 0.00 0.00 0.20 0.05 0.03 0.02 0.01 0.00 0.50 0.13 0.09 0.04 0.00 0.00 1.00 0.25 0.17 0.08 0.00 0.01 5.00 1.23 0.84 0.41 0.04 0.02 10.00 2.40 1.63 0.77 0.08 0.03 20.00 4.56 3.04 1.42 0.15 0.05 NOTE: Exposure rate assumed constant starting at birth.
|
From page 76... ...
The estimated lifetime risk attributed to the NCRP was calculated as follows, from Table 10.2 in NCRP Report 7820 "Lifetime lung cancer risk under environmental conditions per W[M per year." The lifetime risk for lifetime exposure from age 1 ~ 9.1 x 10-3, assuming an average life span in 1976 of 70 yr, yields 9.1 x 10-3/70 WLM _ 1.3 x 10~4 cases per 1 WI,M or 130 per lo6 person-W[M of lifetune exposure. the middle ages.
|
From page 77... ...
SUMMARY AND RECOMMENDATIONS Radon and its daughter products are ubiquitous in indoor environments. Underground miners, exposed to radon daughters in a Here's air, have an increased risk of lung cancer that has been demonstrated in numerous populations.
|
From page 78... ...
For the purpose of risk estimation the committee assumed that the findings in the miners could be extended across the entire life span, that 1 W[M yields an equivalent dose to the respiratory tract in both occupational and environmental settings, that cigarette smoking and radon-daughter exposure interact multiplicatively, and that the sex-specific baseline risk of lung cancer ~ increased multiplicatively by radon daughters for males and females. The committee judged that its assumptions were supported by available evidence, although some alternatives are possible.
|
From page 79... ...
~ Additional information on the interaction between radon daughters and cigarette smoking with regard to the induction of lung cancer is needed. Both animal data and epidem~ological investigations are relevant.
|
From page 80... ...
1981. A model for predicting lung cancer risks induced by environmental levels of radon daughters.
|
From page 81... ...
1979. An exposuretime-response model for lung cancer mortality in uranium miners—ejects of radiation exposure, age, and cigarette smoking.
|
From page 82... ...
1984. Lung cancer in Swedish iron miners exposed to low doses of radon daughters.
|
From page 83... ...
1981. Mortality follow-up through 1977 of the white underground uranium miners cohort examined by the United States Public Health Service.
|
From page 84... ...
It also provides a unified approach for testing the validity of models and for estimating the value of parameters. For the purposes of interest here, a particular strength of the new methods is that they permit the analysis of cohort data with purely internal comparisons, rather than comparison with external population rates, as in traditional SMR methods.
|
From page 85... ...
Lung cancer is caused by many different agents that may cause the disease on their own or through combined effects. The process of carcinogenesis is undoubtedly very complex, and any mode!
|
From page 86... ...
is correct, because the two models are simply alternative expressions of the excess risk. Our analysis places substantial emphasis on comparisons purely internal to the cohorts, in contrast to comparisons with external population rates.
|
From page 87... ...
In statistical terminology, r is essentially what is called a hazard function, but we will refer to it, rather imprecisely, simply as risk. For an individual, T iS thus the age-specific risk, that is, the chance of dying of lung cancer in 1 yr, at age a, given that he is at risk (is still alive)
|
From page 88... ...
Although this assumption can, in principle, be checked by analysis of cohort data, modeling of excess risk becomes a morass of complexity without a tentative assumption of this nature. In particular, without it there would be no concept of a dose effect independent of the values of the other variables.
|
From page 89... ...
In the radioepidemiologic tombless this type of mode} was chosen, after very careful consideration, for describe ing the temporal patterns of radiation-induced cancers other than leukemia. Because the background rates of epithelial cancers generally increase roughly as a power of age, with an exponent of perhaps 5, the absolute excess risk increases very sharply with age as well.
|
From page 90... ...
The simplest and most commonly used approach is to take the dose variable to be cumulative exposure up to the age at risk, except for a few years' lag. The lag allows for a minimal period between the exposure and the expression of the risk, as manifested by diagnosis or death.
|
From page 91... ...
= pk. then this reduces to a mode} of the form in Equation 2A-3, using only cumulative dose to age a (lagged)
|
From page 92... ...
Among smokers, however, the relative risk for dose is now dependent on whether Equation 2A-8 or 2A-9 Is more appropriate. The relationship between smoking and radon exposure Is very complex, and the best description of the association could vary from supramultiplicative to subadditive, although analyses presented in Appendix VIT clearly suggest that the relationship is not additive or subadditive.
|
From page 93... ...
Previously, this methodology appears to have been used for the cohorts of subjects exposed to radon daughters only for the Colorado Plateau data, by Whittemore and McMilian2425 and Hornung and Meinhardt.7 We believe that a thorough analysis of these cohort data required methods of essentially the same statistical nature as those described here. There are many variations on these approaches, for example,
|
From page 94... ...
is fitted consists of a cross-cIassification, in cells defined by specified intervals of a, p, and it, of the numbers of deaths due to lung cancer, the numbers of person-years at risk, and the mean dose for the cell. The intervals of age and calendar period should be fairly narrow, such as 5 yr.
|
From page 95... ...
For each of these cells, then, we record the length of time the individual spent in the cell, whether the individual became a "case" there, and the actual dose for the chosen point in the cell. Finally, the data are accumulated over all individuals by adding up for each cell the times spent and the number of cases and by computing the mean dose for that cell.
|
From page 96... ...
In the regression analysis, the excess relative risk fta) d in Equation 2A-13 can be replaced by various expressions that depend on these factors, so that their effect on the relative risk can be tested.
|
From page 97... ...
A commonly used manner of calculating the dose response is to compute an SMR for each dose category, and then to regress the SMRs in a suitable way on the mean dose levels for each category. The dose-specific SMRs are computed from the ciata summary described above as c(~/b(~)
|
From page 98... ...
mean doses is the same as would be obtained by our relative-risk regression method for the case that ,B(a) in Equation 2A-13 is taken independent of a.
|
From page 99... ...
The aim in this was to examine the dependence of the relative risk on several factors: age, time since exposure, age at first exposure, and exposure rate
|
From page 100... ...
. In particular, analyses were carried out to examine the adequacy of models where the relative risk is taken to be constant for a given cumulative exposure.
|
From page 101... ...
For the Ontario and Colorado miners, we have taken this to be the date of the first annual physical during the study period. Cumulative exposures were computed with a 5-yr lag of current age, to allow for a minimal period when the exposures would not be expected to result in a substantial risk of Jung-cancer mortality.
|
From page 102... ...
The comparison is to external population rates. Figures 2A-1 through 2A-4 show the excess relative risk in exposure categories for each cohort.
|
From page 103... ...
Lung-cancer deaths: 51 Person-years at risk: 27,397 Colorado (January 1, 1951, to December 31, 1982) Total Lung-cancer deaths: 256 Person-years at risk: 73,642 Under 2000 W[M Lung-cancer deaths: 157 Person-years at risk: 66,237
|
From page 104... ...
Although this amount of detail will ordinarily be of interest only after seeing the results of our analysis, it will be convenient to refer to these tables occasionally in what follows. As described in the statistical methods section, the analysis was done in terms of grouped data that define the cells for which dose, cases, person-years, mean doses, for example, are tabulated.
|
From page 105... ...
LLI O 2.0 0.0A -1.0 COLORADO ~ L 0 200 500 1000 1500 2000 2500 1 EXPOSURE (WLM) FIGURE 2A-4 Excess relative risk and 67% confidence limits by exposure category as observed in the uranium miners on the Colorado Plateau.
|
From page 106... ...
106 ~ ~ ~ ~ o _ oo O oO ~ ~ oO ~ r~ _ .
|
From page 107... ...
107 o ~ _ oo ~ ~ ~ o ~o U)
|
From page 108... ...
Tables 2A-2 and 2A-3 summarize the main results from analysis with these types of models. For the Colorado cohort only experience where exposures are less than 2,000 W[M have been used; the rationale for this is fairly apparent from the lack of linearity seen in Figure 2A-4, and will be discussed in more detail below.
|
From page 109... ...
bCumulative exposures >2,000 WLM are excluded. CNumbers in parentheses are multiplicative standard errors.
|
From page 110... ...
bCumulative exposures > 2,000 WLM are excluded. CNumbers in parentheses are multiplicative standard errors.
|
From page 111... ...
As outined above, these statistics are based on likelihood ratio tests and are preferable ~ O ~ ~ ~ ~ O on several grounds to assessing the significance in terms of the variation in risk estimates and their standard errors. For example, in Table 2A-2 for Ontario and variation in relative risk with age, there would be approximately a 30% chance of obtaining this much apparent variation in risk if in fact it were truly constant in age, due simply to inherent random variation in estimation.
|
From page 112... ...
Turning to the effect of duration of exposure, it is emphasized that since the relationship examined here is, in effect, adjusted for cumulative exposure, this could equally be called an effect of exposure rate. The only significant or consistent such effect found here is in the Colorado cohort.
|
From page 113... ...
The inferences from Tables 2A-2 and 2A-3 regarding effects of factors other than cumulative dose are affected very little by fixing this SMR. In the analyses of the combined data sets to obtain our final risk estimates, the Malmberget estunate is given relatively small weight because these data are less extensive than are those for the other cohorts.
|
From page 114... ...
The Colorado cohort does not exhibit the time since exposure effect at under 2,000 WHIM; but, as with all the effects examined here, this aspect is not well-estimated from that cohort alone. Moreover, this effect does emerge when the wider range of exposures there are considered.
|
From page 115... ...
In view of these definitions, it follows that ,8(cohort) is specifically the excess relative risk per 100 WHIM incurred during the period ~10 yr before the current age for one of age 55-64 yr.
|
From page 116... ...
Numbers in parentheses are multiplicative standard errors. edf = Degrees of freedom.
|
From page 117... ...
USE = Time since exposure. eNumbers in parentheses are multiplicative standard errors.
|
From page 118... ...
In the external analysis of the Colorado cohort alone, the risk due to exposures 10-15 yr back is estimated as only 0.8/0.7 of that due to exposures 5-10 yr back (Table 2A-3) , a ratio smaller than 1.3 from the combined analysis.
|
From page 119... ...
It is a routine and general statistical calculation that if four independent estimates of the same quantity are obtained, estimates whose logarithms are normally distributed and whose multiplicative standard errors are 1.5, in line with the estimates here, then there is a 10%0 chance that the ratio of the largest to the smallest will be at least 4.6, which is about the same as the variation found here. Thus, particularly in view of the inevitable systematic differences in dosimetry and other substantial inadequacies in studies such as
|
From page 120... ...
Those analyses, however, include a substantial decrease in relative risk with age at exposure, and this is difficult to distinguish from a decline with age. If models are fitted to the Japanese atomic-bomb survivor data with no (or substantially less)
|
From page 121... ...
Refitting the data with the three parameters fixed at these values yields for ,B, the excess relative risk per 100 W[M cohort-specific risk for the baseline categories of age and time as: Internal External Cohort Analysis Analysis Eldorado 5.1 5.0 Ontario 1.8 2.0 Malmberget 3.6 3.7 Colorado 0.9 0.7 When the combined data are fitted by replacing these cohort-specific parameters by a common parameter for ad cohorts, the estimate of IS for the intern e] analysis is 2.2 and for the external analysis is 2.6.
|
From page 122... ...
The multiplicative standard errors for these estimates are 1.3 and 1.2.
|
From page 123... ...
TABLE 2A-6b Observed and Expected Lung Cancer Deaths and Excess Relative Riska Cross-Classified by Age at First Exposure and Cumulative Dose for the Eldorado Cohort Age at First Exposure (yr)
|
From page 124... ...
- 11 bExcess relative risks computed to be less than zero are reported as zero. TABLE 2A-6d Observed and Expected Lung Cancer Deaths and Excess Relative Riska Cross-Classified by Duration of Exposure and Cumulative Dose for the Eldorado Cohort Duration of Exposure (yr)
|
From page 125... ...
- 11 hExcess relative risks computed to be less than zero are reported as zero. TABLE 2A-7b Observed and Expected Lung Cancer Deaths and Excess Relative Riska Cross-Classified by Age at First Exposure and Cumulative Dose for Ontario Uranium Miners Age at First Exposure (yr)
|
From page 126... ...
TABLE 2A-7d Observed and Expected Lung Cancer Deaths and Excess . Relative Riska Cross-Classified by Duration of Exposure and Cumulative Dose for Ontario Uranium Miners Duration of Exposure (yr)
|
From page 127... ...
TABLE 2A-8b Observed and Expected Lung Cancer Deaths and Excess Relative Riska Cross-Classified by Age at First Exposure and Cumulative Dose for the Malmberget Cohort Age at First Exposure (yr)
|
From page 128... ...
TABLE 2A-8d Observed and Expected Lung Cancer Deaths and Excess Relative Riska Cross-Classified by Duration of Exposure and Cumulative Dose for the Malmberget Cohort Duration of Exposure (yr)
|
From page 129... ...
- 11 h excess relative risks computed to be less than 0 are reported as 0. TABLE 2A-9b Observed and Expected Lung Cancer Deaths and Excess Relative Riska Cross-Classified by Age at First Exposure and Cumulative Dose for the Colorado Cohort Age at First Exposure (yr)
|
From page 130... ...
TABLE 2A-9d Observed and Expected Lung Cancer Deaths and Excess Relative Riska Cross-Classified by Duration of Exposure and Cumulative Dose for the Colorado Cohort Duration of Exposure (yr)
|
From page 131... ...
To calculate the probability of dying of lung cancer, suppose qi is the probability of surviving year i when all causes are acting, conditional on surviving through year i- 1; h`* is the mortality rate due to all causes; and hi is the lung-cancer mortality rate at year i.
|
From page 132... ...
= 0.9984. The additional risk of lung cancer due to exposure to radon progeny is incorporated into these risk calculations through the age-specific lung-cancer mortality rates.
|
From page 133... ...
THE COMMITTEE'S ANALYSIS OF FOUR COHORTS OF MINERS 133 ~ to o V .~ ~ 3 ~ o ~ ~ 3 4 is C
|
From page 134... ...
= 0.7877. The conditional lifetime risk of lung cancer (between 60 and 110 yr)
|
From page 135... ...
1981. A model for predicting lung cancer risks induced by environmental levels of radon daughters.
|
From page 136... ...
1984. Lung cancer in Swedish iron miners exposed to low doses of radon daughters.
|
From page 137... ...
ANNEX 2B Radon Dosi~netry This annex gives background information about the radon-222 decay chain and its dosimetry, the entry and deposition of radon daughters in human lungs, and the factors influencing dose per unit exposure In underground mines and homes. DECAY OF RADON AND ITS DAUGHTERS The decay scheme of radon-222, starting with its parent isotope radium-226 and ending with stable lead-206, is shown in Figure 2-!
|
From page 139... ...
If the radon progenies are produced in air, most of the charged atoms rapidly become attached to aerosol particles. Because the proportion of ions that do not become so attached is particularly important in the dosimetry of radon progeny, it has been given a special designation, that is, the unattached fraction f.
|
From page 140... ...
is a collection of mathematical functions used to calculate absorbed doses. The dose delivered by radon daughters to the lung depends on both the aerosol involved and the physiology of the lung.
|
From page 141... ...
This section describes some of the recent dosimetric models for radon daughters. In 1964, Altshuler et al.t and Jacobi2t introduced the first modern lung models which described the inhalation, deposition, and retention of the radon daughters.
|
From page 142... ...
Present dosimetry models assume that every radionuclide of a given type has the same probability per unit time for removal from a compartment (i.e., first-order kinetics) such as those shown in Figure 2~1.
|
From page 143... ...
The first step for the mathematical modeling of the respiratory tract is compartmentalization, as shown in Figure 2B-1. The dimensions and relative orientations of the airways these compartments represent are based on measurements of human lungs.47 43 44 To oh tain a computer-manageable structure for the respiratory tract, the measurements are averaged to give a symmetrical branching structure; that is, each division of an airway was assumed to give two new airways of equal size at the same angle to the original airway.
|
From page 144... ...
The differences between the models for different particle sizes, unattached fractions, and minute volumes will be seen to be only a factor of 2 or 3 within the normal range of these factors. A comparison of the models gives a measure of one source of uncertainty in the determination of dose from radon daughters.
|
From page 145... ...
In most circumstances, inhaled aerosob have a fairly wide distribution of sizes. For aerosols with attached radon daughters in mines, George et al.9 measured activity median diameters of 0.1 to 0.3 ,um with geometric standard deviations as of about 2.7.
|
From page 146... ...
Each of the three models treats the unattached fraction differently. Figure 2~3 shows how the mean dose per unit exposure varies as a function of percent unattached fraction in each model.
|
From page 147... ...
The entry of radioactivity into the lung depends on the amount of air inhaled per unit time (the minute volume) , which is the product of two quantities: the tidal volume, the volume of air inhaled per breath, and the breathing rate, the number of breaths per unit time.
|
From page 148... ...
Recently, James27 has reported that the total deposition is proportional to the square root of the minute volume over the range of particle sizes in which deposition is due to particle diffusion. The variation in the mean bronchial dose as a function of minute volume for the three models is shown in Figure 2~4 for particles with an AMAD of 0.15 ,um.
|
From page 149... ...
. In addition to differences in the loci of the bronchi for which the dose was calculated, differences in input values of the parameters describing the radon daughters in the air and of the parameters characterizing the dosimetry models were responsible for much of the spread.
|
From page 150... ...
Rather than postulating some average or typical values for the various factors that affect the dosimetry of radon progeny in mines and homes and calculating appropriate dose per WI,M conversion factors, the committee proposes a methodology that can be applied when any radon environment is sufficiently characterized. By using the ratio of the dose per W[M in nones to that in homes obtained by means of the lung models described above, it is possible to extrapolate the risk values obtained for underground miners to people living in homes, at a reasonable level of approximation.
|
From page 151... ...
It is known that clearance in the elderly is delayed.~° Most lung models show an increase of bronchial dose per unit exposure until about age 6, after which it falls off and becomes nearly constant after age 10. The NEA Organisation for Economic Co-operation and Development has recommended that age dependency of dose equivalent per unit exposure be neglected.35 It is instructive to examine the variations of the proportionality factor K, given by Equation 2B-4, under the various dosimetry models, as illustrated by the results shown in Figures 2B-2, 2B-3, and 2~4.
|
From page 152... ...
For example, with increasing aerosol concentration the dose per W[M conversion factor for equilibrium (F) generally increases somewhat, while that for the unattached fraction (0 decreases.
|
From page 153... ...
of about 4%, the three models shown in Figure 2B-4 give a conversion factor to the bronchial region of about 0.5 0.6 rad/W~M. For a diffusion coefficient of 0.005 cm2/s and an unattached fraction of To in homes, the same curves give conversion factors between 0.5 and 0.7 rad/W[M.
|
From page 154... ...
ICRP,~9 NEA,35 and NCRP33 assign a breathing rate of 20 liters/m~n to underground miners and base their dose estimates on this minute volume. Ventilation rates in working coal miners have been measure; values obtained ranged from about 30 to 40 liter/min when averaged throughout the working shift.
|
From page 155... ...
On the other hand, if the minute volume for miners is taken as 20 liters/min, the ratio K is increased to about 1.3. In homes where the unattached fraction is low (Rio)
|
From page 156... ...
1969. Deposition of radon daughters in humans exposed to uranium mine atmospheres.
|
From page 157... ...
1986. Dosimetric approaches to risk assessment for indoor exposure to radon daughters.
|
From page 158... ...
1985. Measurements of unattached fractions of radon daughters in houses.
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.