3.
ESTIMATION OF RADIATION RISKS OF
THYROID CANCER IN THE POPULATION
OF THE ORYOL OBLAST
There is evidence that exposure of thyroid to ionizing
radiation results in an increase of thyroid cancer incidence. The magnitude of radiation risk, that is the probability
of thyroid cancer occurrence, is high and comparable to the risk of leukemia.
The risk of thyroid cancer depends on age at exposure
and increases with decreasing age. Given thyroid exposure to
incorporated iodine isotopes, the dose is also a function of age at
exposure and increases with decreasing age at exposure. The consequences of
exposure show themselves as an increase in thyroid cancer incidence rate after
a latent period of at least 4 years.
One of possible causes of the increase in thyroid
cancer incidence in the Oryol oblast may be associated with thyroid exposure to
incorporated iodine isotopes. The relationship of increase in thyroid cancer
incidence and exposure to incorporated iodine isotopes after the
A General description of the
population being studied
The analysis of radiation risks and of the dose
response relationship of thyroid cancer incidence focuses on the population of
the Oryol oblast represented by residents of 3006 settlements. The demographic
characteristics of the population are those obtained during the 1989 census and
include the size of age groups by rayons and the number of people living in
separate settlements. Adequate assessment of thyroid doses from incorporated ^{131}I
for cases and the population in general requires more details at the level of
settlement. Since children and adolescents (at time of exposure) form a group
who have an increased radiation risk, this group was considered separately.
This group consists of members of the population born between 1969 and 1986.
The size of the whole population of the Oryol oblast in 1989 was 874046
persons, of them 207,624 were children and adolescents (017 years) and 511,716
were adults (1860 years). Excluding the city of
The followup period for the studied population covers
1991 to 2001. The time of the beginning of followup 1991 was selected with
allowance for minimal latent period of radiogenic cancers of 5 years.
Cases
The available data on thyroid cancer cases in the
Oryol oblast are official findings of the Oryol oncological clinic. The
diagnosed cases between 1991 and 2001 for those exposed as children and
adolescents and adults is shown in Figs. 3.1 and 3.2. Figures 3.3 and 3.4
present the cases for the population excluding the city of Oryol.
As can be seen from Figs. 3.13.4, the peak incidence
occurred between 1996 and 1998. A total of 78 thyroid cancers in children and
adolescents at exposure were diagnosed and 777 cases among adults between 1991
and 2001. With exclusion of Oryol, the number of cases in children and
adolescents is 34 and 338 cases of thyroid cancers among adults.
Fig.
3.1. The thyroid cancer cases in those who were children and adolescents at
time of exposure.
Fig.
3.2. The thyroid cancer cases in adults (1860 years of age at exposure).
Fig.
3.3. The thyroid cancer cases in those who were children and adolescents at
exposure
(excluding the city of Oryol).
Fig.
3.4. The thyroid cancer cases in those who were adults at exposure
(excluding the city of Oryol).
Table 3.1 contains the size of the populations and the
number of cases for the whole population (060 year of age at exposure) in
major towns (Oryol, Livny and Mtsensk) and rayons of the Oryol oblast. It
follows from Table 3.1 that 486 cases, which make more than half of all cases,
have been identified in the city of
Since age at exposure is a factor of risk, Figs.
3.53.8 show distribution of cancer cases as a function of age at exposure.
Table 3.1. Number of
cases over the followup period 19912001 by rayons and major towns.
Rayon 
Korsakovsky 
Bolkhovsky 
Verkhovsky 
Glazunovsky 
Dmitrovsky 
Population 
4196 
17390 
19278 
13782 
13072 
Number of
cases 
3 
42 
15 
10 
6 
Incidence
rate per 100000 persons 
53.8 
182.1 
60.9 
58.8 
32.8 
Mean dose,
mSv 
17.0 
29.5 
12.7 
22.3 
33.1 
Rayon 
Dolzhansky 
Zalegoshchensky 
Znamensky 
Kolpnyansky 
Krasnoozerensky 
Population 
11748 
14392 
4808 
17289 
6956 
Number of
cases 
5 
14 
3 
10 
3 
Incidence
rate per 100000 persons 
32.4 
74.4 
45.6 
44.4 
31.7 
Mean dose,
mSv 
7.8 
14.0 
12.8 
11.4 
16.3 
Rayon 
Kromsky 
Livensky 
Maloarkhangelsky 
Mtsensky 
Novoderevenkovsky 
Population 
18888 
26861 
12236 
15889 
11652 
Number of
cases 
23 
10 
6 
18 
1 
Incidence
rate per 100000 persons 
92.3 
29.0 
38.2 
80.1 
6.8 
Mean dose,
mSv 
22.6 
11.9 
30.0 
15.4 
13.6 
Continuation of Table 3.1.
Rayon 
Novosilsky 
Oplovsky 
Pokrovsky 
Sverdlovsky 
Soskovsky 
Population 
8831 
50340 
15105 
15109 
6689 
Number of
cases 
2 
55 
8 
16 
2 
Incidence
rate per 100000 persons 
11.8 
86.6 
39.7 
82.2 
21.7 
Mean dose,
mSv 
15.7 
13.8 
15.2 
21.1 
17.5 
Rayon 
Trosnyansky 
Uritsky 
Khotynetsky 
Shablykinsky 
Oryol 
Population 
10432 
14400 
9949 
8265 
287695 
Number of
cases 
10 
15 
9 
5 
486 
Incidence
rate per 100000 persons 
70.3 
78.1 
69.0 
46.5 
145.9 
Mean dose,
mSv 
21.5 
16.6 
10.2 
15.6 
16.1 
Rayon 
Livny 
Mtsensk 
Population 
45265 
42919 
Number of
cases 
43 
38 
Incidence
rate per 100000 persons 
83.1 
78.0 
Mean dose,
mSv 
7.7 
12.6 
Fig.
3.5. Distribution of cases in children and adolescents by age at exposure.


Fig.
3.6. Distributions of cases in adults by age at exposure and age
distribution
of the population of the Oryol oblast (060 years of age).
Fig.
3.7. Distribution of cases in children and adolescents by age at exposure
(excluding the city of
Fig.
3.8. Distribution of cases in adults by age at exposure
(excluding the city of
For those exposed as children and adolescents the
incidence increases with age, which is consistent with the age dependence of
spontaneous incidence. However, it is worth pointing to the increase in the
number of cases at 1 year of age (the age when the radiation risk is maximal).
For adults the peak incidence occurs at the age of 3537 and 4550 years. In
those more than 50 years of age the number of cases decreases. The gap in the
age distribution for persons between the ages of 4045 years is explained by
the demographic gap in the age distribution of the population (the right part
of Fig. 3.6).
Radiation doses
As was mentioned in Chapter 1, thyroid doses have been
reconstructed using the official guidelines adopted by the Russian Scientific
Committee on Radiation Protection. The mean doses by rayons of the Oryol oblast
for the studied group (the ages of 060 years at exposure) are included in
Table 3.1. As can be seen from the table, the largest dose values of about 30
mSv occurred in the Bolkhovsky, Dmitrovsky and Maloarkhangelsky rayons. Table
3.2 includes the mean thyroid dose values (mSv) for different age groups.
Table
3.2. Mean thyroid doses (mSv) for different age groups in the Oryol oblast
(persons exposed at different age).
Age group 
017 
1860 
Oblast as a
whole 
36.2 
7.6 
Excluding
the city of 
31.4 
7.8 
It follows from Table 3.2 that the mean dose for those
exposed as children and adolescents is about 5 times higher that in adults.
Technique of analysis of the dose
response relationship
of thyroid cancer incidence
The analysis was performed using modern approaches.
For descriptive analysis the standardized incidence ratio (SIR) is used equal
to the ratio of the observed cases to the number of cases expected with no
exposure. SIR is estimated using the indirect standardization method usually
applied for rare diseases such as leukemias and thyroid cancers.
Standardization of an indicator makes it possible to take into account
differences in the age structure of compared groups. For rare diseases
nationwide agespecific indicators are normally used as a control. In this
study the controls were agespecific thyroid cancer incidence rates for Russia
during the time period from 1991 to 2002.
For calculation of SIR and 95% confidence intervals
the statistical package EPICURE [7] developed specifically for study of health
effects of exposure in the cohort of atomic bomb survivors in Japan. The SIR
was calculated using the formula:
_{},
where summation was made by time (index j) and age groups
(index i); cases_{i,j} is the
number of cases in the age group (i) at the time moment (j);
PY_{i,j} is the number of personyears of followup in
the age group (i) at the time moment (j); _{} are agespecific thyroid cancer incidence
rates in Russia in the age group (i) at the time moment (j).
The radiation risks and dose response parameters were
calculated using the EPICURE package (the module AMFIT estimating regression
coefficients by the maximum likelihood technique for grouped and stratified
data).
For consideration of age and time differences in the
compared groups the data were stratified by attained age and calendar time and
divided into 4 dose groups: 0, 0.012, 0.02, 0.05, >0.05 for children and
adolescents and 0, 0.006, 0.007, 0.0010, >0.01 Sv for adults.
The calculations were made for models with internal
and external controls. The risk model with an internal control takes the form:
_{},
where l_{i,j} is the thyroid cancer incidence rate in the stratum (i,j);
_{} is the spontaneous incidence rate in the Oryol
oblast in the stratum (i,j); ERR_{1Sv} is the
excess relative risk at the dose of 1 Sv; d_{i,j} is the
thyroid dose in the stratum (i,j).
The model with an external control is described by the
equation:
_{},
where SIR^{un} is the standardized incidence ratio
for unexposed population.
The confidence (95%) intervals were estimated by the
likelihood function profile. In addition, for analysis of the dose response relationship,
the distribution of standardized incidence ratio (SIR) was derived. This can be
considered as an approximate distribution of relative risk by radiation doses.
Since the method of thyroid dose reconstruction and the dose values
themselves are still debatable, the effect of radiation exposure on thyroid
cancer incidence was studied using the same technique as in the analysis of the
Bryansk oblast data [8]. This approach allows qualitative assessment of the
role of the radiation exposure in thyroid cancer incidence without using dose
values. The gist of the method is as follows. The radiation risk of thyroid
cancer per unit dose and thyroid dose tends to increase with decreasing age at
exposure. This means that exposure effects should be seen among children of
smaller age. For ascertaining it is sufficient to draw a distribution of the
population and cases by age at exposure as a factor of risk. The frequency of
spontaneous cases, given no exposure, should be proportional to the size of a
particular age group and the distributions should coincide. If the radiation
response makes itself evident in some age groups, then the distribution of
cases and that of the population in general will be different. The workability
of the method was demonstrated in the study of thyroid cancer incidence in the
Main results
Children and adolescents at exposure (017 years
old)
The dynamics of SIR for children and adolescents at
exposure is shown in Figs. 3.9, 3.10. As can be seen from the figures, the SIR
reaches its maximum in 19941996 exceeding the nationwide incidence level by a
factor of 67 and then decreases to the control level (values 12). Such a time
dependence can be attributed either to screening effect (detection of earlier
diseases) or early manifestation of exposure effects in this category of the
population.
Fig.
3.9. The standardized incidence ratio (SIR) for children and adolescents at
exposure.
Fig.
3.10. The standardized incidence ratio (SIR) for children
and adolescents at exposure (excluding the city of
Let us now consider the distribution of cases and the
population as a whole by age at exposure (Figs. 3.11 and 3.12). As can be seen
from Fig. 3.11 for the entire population, the distribution patterns are almost
identical, which in a way suggests an absence of an exposure effect. Excluding
Oryol from the study, the distribution of cases is shifting towards younger
ages, which may be an indication of the exposure factor, however, to a lesser
extent than in those exposed as children and adolescents living in the Bryansk
oblast (Fig. 3.13).
The SIR dose response relationship is presented in
Figs. 3.14 and 3.15. It can be seen that the SIR dependence has a slight
positive trend for thyroid dose, both for the population as a whole and with
exclusion of Oryol.
Fig.
3.11. Distribution of the population and cases by age at exposure.
Fig.
3.12. Distribution function of the studied population and cases
(excluding Oryol) by age at exposure.
Fig.
3.13. Distribution of the population and the population as a whole
by age at exposure (girls of the
Fig.
3.14. Standardized incidence ratio (SIR) as a function of dose.
Fig.
3.15. Standardized incidence ratio (SIR) as a function of dose
(excluding Oryol).
Adults of 1860 years of age at
exposure
Figs. 3.16 and 3.17 present the SIR for adults. As can
be seen from the figures, SIR has similar features for adults to those in
children and adolescents: an increase in the initial followup period, the peak
in 19961998 and a decrease towards the control level at the end of the
followup period.
The dependence of SIR on dose for adults is shown in
Figs. 3.18 and 3.19. As follows from the figures, the dependence has a slight
positive trend, i.e. the incidence increases with increasing radiation dose.
Results of calculating the standardized incidence
ratio with 95% confidence intervals over the whole followup period from 1991
to 2001 for different age groups are shown in Table 3.3.
Fig.
3.16. The standardized incidence ratio (SIR) for adults at exposure.
Fig.
3.17. The standardized incidence ratio (SIR) for adults at exposure
(excluding Oryol).
Fig.
3.18. Standardized incidence ratio (SIR) as a function of dose.
Fig.
3.19. Standardized incidence ratio (SIR) as a function of dose
(excluding Oryol).
Table 3.3. Results of
calculating the standardized incidence ratio over the followup period from
1991 to 2001 for different age groups.
Population
group 
Children
and adolescents (017) 
Adults
(1860) 
Whole
population 
4.04 (3.20, 5.03 95% CI) 
2.42 (2.25, 2.60 95% CI) 
Excluding
Oryol 
2.88 (1.98, 4.04 95% CI) 
1.72 (1.55, 1.92 95% CI) 
It follows from the Table 3.3 that incidence in
children and adolescents at exposure living in the Oryol oblast in the
followup period is 34 times higher than the nationwide indicators. For adults
this excess is a factor of 1.72.5.
A question about the influence of radiation exposure
on incidence can be better answered based on the regression analysis of the
dose relationship of incidence. For this purpose, as was mentioned above, the
statistical package EPICURE was used.
In calculations with EPICURE the data were grouped
into 4 dose groups. The key parameters of the dose groups for the whole
population are illustrated in Table 3.4.
Estimates of radiation risk coefficients (excess
relative risk per unit dose 1 Sv (ERR_{1Sv})) calculated with EPICURE
are presented in Table 3.5.
Table 3.4. Key
parameters of dose groups for the whole population.
Dose groups 
Mean dose
(mSv) 
PY (person years) 
Cases 
Spontaneous
cases* 
Mean dose
(mSv) 
PY (person years) 
Cases 
Spontaneous
cases 
Children
and adolescents
(017) 
Adults (1860) 

1 
0.9 
487100 
27 
25.6 
0.4 
1370510 
141 
183 
2 
15.5 
514512 
25 
20.0 
6.5 
580565 
55 
77 
3 
31.8 
757000 
16 
15.6 
7.4 
2851080 
488 
373 
4 
88.1 
524147 
10 
5.5 
14.4 
826425 
93 
108 
Total 
36.2 
2282759 
78 
66.7 
7.6 
5628080 
777 
741 
* estimate of spontaneous cases was
derived from calculations with EPICURE with an internal control.
Table 3.5. Estimates
excess relative risk per unit dose 1 Sv (ERR_{1Sv}).
Population
category 
Children
and adolescents
(017) 
Adults (1860) 
Internal
control 

Whole population 
7.8 (n.d., 124.7 95% CI)* 
6.3 (16.7, 43.7 95% CI) 
Excluding Oryol 
13.5 (n.d., 296.3 95% CI) 
12.9 (12.8, 55.2 95% CI) 
External
control 

Whole population 
0.8 (6.4, 16.9 95% CI) 
1.1 (19.9, 34.5 95% CI) 
Excluding Oryol 
9.9 (5.1, 65.0 95% CI) 
9.4 (14.8, 49.0 95% CI) 
*n.d.  the lower confidence limit has not been defined.
It can be seen from Table 3.5, that the values of the
radiation risk coefficients are positive both for children and adolescents and
adults, but are not statistically significant (the lower confidence limit is
negative). In two calculations, the confidence limit was not determined for
children and adolescents because of the shapes of the likelihood function (no
convergence in solution of the system of likelihood equations). It was an
unexpected result that the trends for adults were positive. In the studies for
the Bryansk oblast the trend for adults was negative.
Descriptive
statistical analysis of the relationship of the thyroid cancer incidence
in the Oryol oblast and the radioactive ^{131}I contamination of soil
and
the stable iodine content in soil
This section provides results of a descriptive
analysis of the dependence of thyroid cancer incidence in the Oryol oblast on
two environmental factors: the radioactive ^{131}I contamination of
soil and the stable iodine content in soil. Each factor was represented by 6
levels of concern, as can be seen from Table 3.6.
Two followup periods were studied: from 1982 to 1991
assuming that no radiationinduced cancers occurred during this period (the
latent period was from 1986 to 1991) and the time between 1992 and 2001.
Table 3.6. Levels of
concern of environmental factors.
Levels of
factors 
Level of
iodine in soil 
Radioactive
^{131}I in soil, kBq/m^{2} 
1 
Extremely low 
1020 
2 
Very low 
2040 
3 
Low 
40100 
4 
Moderately low 
100300 
5 
Lower limit of norm (forested area) 
300600 
6 
Lower limit of norm (foreststeppe) 
6001200 
The descriptive analysis is represented by the values
of thyroid cancer incidence rate (95% confidence intervals) as a function of
the mentioned environmental factor levels. The source data were the rayon
incidence rates and each rayon was related to levels of stable and radioactive
iodine in soil in accordance with Table 3.6. Results of calculation of the
incidence rate for both genders in different followup periods are shown in
Table 3.7. As can be seen from Table 3.7, the population of the Oryol oblast is
mainly concentrated in the territories with levels 46 with respect to stable
iodine and levels 23 with respect to contamination with radioactive iodine.
The influence of iodine levels on incidence was
studied for persons exposed between the ages of 060 years. Children and
adolescents were not considered separately, as the number of cases for them is
limited.
The data of Tables 3.7 and 3.8 are presented in graphs
of Figs. 3.20 and 3.21. The indicators for the followup period between 1992
and 2001 are noticeably (34 times) above those for the period from 1982 to
1991.
Table 3.7. Calculated
incidence rates for both genders (followup period 19821991).
Level 
Number of
cases 
Followup
person years 
Rate per
100000 persons 
Stable
iodine in soil 

1 
 
 
 
2 
 
 
 
3 
 
 
 
4 
9 
297126 
3.0 
5 
51 
1039504 
4.9 
6 
176 
4968586 
3.5 
Radioactive
^{131}I 

1 
4 
117486 
3.4 
2 
32 
884882 
3.6 
3 
195 
5128947 
3.8 
4 
5 
173901 
2.9 
5 
 
 
 
6 
 
 
 
Table 3.8. Calculated
incidence rates for both genders (followup period 19922001).
Level 
Number of
cases 
Followup
person years 
Rate per
100000 persons 
Stable
iodine in soil 

1 
 
 
 
2 
 
 
 
3 
 
 
 
4 
19 
297126 
6.4 
5 
160 
1039504 
15.4 
6 
648 
4968586 
13.0 
Radioactive
^{131}I 

1 
5 
117486 
4.2 
2 
81 
884802 
9.2 
3 
701 
5128947 
13.7 
4 
5 
173901 
23.0 
5 
 
 
 
6 
 
 
 
Fig.
3.20. Dependence of the thyroid cancer incidence rate on the level
of stable iodine in soil in different followup periods.
Fig.
3.21. Dependence of the thyroid cancer incidence rate on the level
of radioactive ^{131}I iodine in soil in different followup periods.
The incidence rate in the period 1982 to 1991 remained
almost unchanged with variations in the levels of radioactive iodine and stable
iodine (Figs. 3.20 and 3.21).
In the period from 1992 to 1999 an increase in the
levels of stable and radioactive iodine results in increased incidence rates
(Figs. 3.20, 3.21). The positive correlation of thyroid cancer incidence and
radioactive iodine level may suggest the induction of radiogenic cancers. On
the other hand, the increase in incidence rate with increasing level of stable
iodine in soil does not fit the hypothesis that the endemia of stable iodine
during fallout of radioactive iodine may be contributing to increase in
radiogenic cancers. The result we have arrived at may be explained by
insufficient accuracy of stable iodine data and by the fact that the iodine
ratio in soil of different territories and in people living there may be
different.
For better visualization of analysis results the
incidence rates (per 100,000 persons) were presented as a matrix (I´J), where I (rows) is the number of levels of stable iodine and J
(columns) is the number of levels of radioactive ^{131}I. Results of
analysis have been differentiated by followup periods, oblasts and sex.
An example of a matrix of thyroid cancer incidence
rate (both genders, 060 years of age at exposure and followup period from
1992 to 1999) is provided by Table 3.9.
Table 3.9. Example of a
matrix of thyroid cancer incidence rates.

0 
1 
2 
3 
4 
5 
0 
0 
0 
0 
0 
0 
0 
1 
0 
0 
0 
0 
0 
0 
2 
0 
0 
0 
0 
0 
0 
3 
0 
7.7 
4.3 
0 
0 
0 
4 
0 
0 
13.9 
23.0 
0 
0 
5 
4.3 
9.5 
13.9 
0 
0 
0 
The cell with maximum incidence rate is shaded.
The matrix was presented as a map with outlines (the
upper lefthand corner of the matrix is in the lefthand corner of the
picture). The red color in the map corresponds to the local maxima of incidence
rates. The abscissa axis is stable iodine level and the axis of ordinates is
radioactive iodine level. Higher number of level means higher iodine content.
Maps with outlines are shown in Fig. 3.22. As can be
seen from the figure, these maps reflect the dependence of incidence rate on
level of stable and radioactive iodine. They also show the values of incidence
rates per 100,000 persons as a function of radioactive and stable iodine level.
If stable and radioactive iodine levels do influence thyroid cancer incidence,
the peak incidence should shift towards the upper lefthand corner, the region
of decreased content of stable iodine and increased content of radioactive
iodine. Figure 3.22 shows a shift in the peak towards higher content of
radioactive iodine.
For quantifying the influence of the considered
factors on thyroid cancer incidence we used the chisquare statistic based on
analysis of differences between expected and observed number of cases [10]. Six
levels of each factor were convoluted into 2 and the data took the form of
table 2´2. The convolution of levels was done by the scheme shown in Table 3.10.
Table 3.10 is a matrix of levels of stable (rows) and radioactive iodine. For
example, cell (5, 4) represents the 5^{th} level of stable iodine and 4^{th}
level of radioactive iodine. Table 2´2 and levels included in the cells
of table 2´2 are boldfaced and delineated. Other cells (not delineated) were not
considered, because such levels of stable and radioactive iodine do not
occurein the Oryol oblast.
Whole
population (060 years) 19821991 
Whole
population (060 years) 19921999 
Fig.
3.22. Maps with outlines of thyroid cancer incidence rate as a function of
level
and a combination of factors.
Table 3.10. Matrix of
the levels of stable (rows) and radioactive iodine.
1.1 
1.2 
1.3 
1.4 
1.5 
1.6 
2.1 
2.2 
2.3 
2.4 
2.5 
2.6 
3.1 
3.2 
3.3 
3.4 
3.5 
3.6 
4.1 
4.2 
4.3 
4.4 
4.5 
4.6 
5.1 
5.2 
5.4 
5.4 
5.5 
5.6 
6.1 
6.2 
6.4 
6.4 
6.5 
6.6 
Comments to the delineated cells of table 2´2 differentiated by the level of influence of factors of stable and
radioactive iodine are illustrated by Table 3.11. The cell “control” means that
the stable iodine content is close to norm and the radioactive iodine
contamination is minimum.
Table 3.11.
Level 
0 
1 
0 
Stable iodine deficiency 
Radiation exposure + iodine deficiency 
1 
Control 
Radiation exposure 
Results of testing of the null hypothesis (the
relative risk is 1) are shown in Table 3.12. The expected number of cases for
the criterion hisquare is calculated by the incidence rate in the cell 
“control’ of table 2´2 and the number of person years in
other cells of table 2´2.
Table 3.12.
Period 
Both genders,
age 060 at exposure 

19821991 
0.98 (a) 
0.33 (c) 
0.90 (b) 
0.93 (d) 

19921999 
0.75 (e) 
<0.001 (g) 
0.93 (f) 
<0.001 (h) 
The cells in which the zero hypothesis was rejected
are shaded. As can be seen from Table 3.12, the incidence in the period 19821991
in the territories with stable iodine (cell d) level close to norm, is not
significantly different from the control (cell b). There is no statistically
significant difference for incidence in the territories with increased density
of radioactive iodine contamination either (cell c). However, in the period
from 1992 to 2001 in the territories in which the influence of stable and
radioactive iodine should be pronounced most of all, the incidence rate (cell
g) is statistically significantly different from the control. A statistically
significant difference from control was also observed for territories with the
content of stable iodine close to the norm and with increased content of
radioactive iodine (cell h) The verification of the hypothesis about the
difference in incidence in cells (h) and (g) has revealed no statistical
significance (p=0.8), i.e. no endemia effect on thyroid cancer
incidence. The fact that no effect of iodine deficiency was observed is
probably explained by a relatively small difference in the levels of stable
iodine in control groups and compared territories (only three categories for
stable iodine 46) and in addition, as was mentioned above, the iodine content
in soil may differ from that in human body. Therefore the presented results
should be considered as preliminary.
Discussion
of the results
The analysis of thyroid cancer incidence showed a
noticeable increase in the incidence in the Oryol oblast as compared with that
for Russia in general. On average, the thyroid cancer incidence in the
population of the Oryol oblast in the time period 1991 to 2001 was 34 times
higher than the nationwide rates, while for adults this excess was a factor of
1.72.5. Possible reasons for this increase can be both regional differences in
incidence rates and manifestation of exposure effects. The SIR since 1991
increases to the maximum values in 19961998, the values being 67 for children
and adults at exposure and 33.5 for adults and then decreases to the levels
close to nationwide rates. Such dynamics of the SIR can be explained by the
screening effect during the studied period (diagnosis of latent diseases as a
result of better registration), manifestation of the radiation response factor
in this time period or a combination of both.
Estimates of the radiation risk reveal a positive
trend of incidence rate as a function of thyroid dose, which, however, is not
statistically significant. A positive trend has been observed for adults as
well. An indirect confirmation to this trend is provided by the distribution of
population and cases by age at exposure (Fig. 3.23) suggesting that the
distributions differ and age of cases shifts to the region of smaller values.
As was said in the section “Methods of analysis”, this may be a qualitative
indication of the influence of radiation factor. For the adults of the Bryansk
oblast these distributions were identical and the incidence trend with
allowance for dose was even negative among adults.
Fig.
3.23. Distribution of adults by age at exposure (excluding Oryol).
In general, the radiation risk factors in the Oryol
oblast are lower than those in the Bryansk oblast. For example, in the Bryansk
oblast the excess relative risk for children and adolescents at exposure with 1
Sv dose was 40 and statistically significant. A similar value was derived for
children and adolescents of
To answer this question let us estimate the
attributive risk which by definition is a proportion of radiogenic risks among all
cases. For this purpose we use data of Table 3.4 showing spontaneous and
observed cases. For children and adolescents the attributive risk is (7866.7)´100/78=14.5% and for adults (777741)´100/777=4.6%. This means that if
estimates of radiation risk derived for the residents of the Oryol oblast are
objective, 15% of thyroid cancers in children and adolescents at exposure are
radiationinduced and 5% for adults.
Assuming that the estimates adopted by UNSCEAR are
objective, the attributive risk for children and adolescents will be equal to
(40´0.036)´100/1+40´0.036=59%.
We shall regard this estimate as conservative, with a
margin of safety.
The performed analysis of the influence of the levels
of stable iodine in soil and radioactive iodine on the incidence shows that
there is a correlation between the incidence and the radioactive iodine level
and no correlation with the stable iodine level. Generally, this relationship
is in agreement with results of radiation risk analysis (positive trend with
dose).
The absence of an effect of iodine deficiency is
probably explained by a relatively small difference in the levels of stable
iodine in control and compared territories (only three categories for stable
iodine 46) and in addition, iodine levels in soil can differ from those in
human body. Therefore, the presented results regarding the influence of endemia
of stable iodine should be treated as preliminary.
Conclusions
1. The analysis has revealed a marked increase in the thyroid cancer
incidence in the Oryol oblast as compared to that for Russia as a whole. On the
average, the thyroid cancer incidence in the Oryol oblast in the period
19912001 was 34 times higher than the nationwide rate, for adults the excess
is a factor of 1.72.5. This increase may be explained by regional differences
in incidence and a manifestation of the exposure effect. The values of the standardized
incidence ratio SIR was increasing from 1991 and reached its maximum in
19961998: the value was 67 for children and adolescents at exposure and 33.5
for adults. Then it was decreasing to the levels close to nationwide indictors.
Such dynamics of the SIR can probably be attributed to the screening effect in
the considered period (diagnosis of latent diseases due to better
registration), the effect of radiation exposure or a combination of these
factors.
2. Estimates of radiation risk show a positive trend of incidence as a
function of thyroid dose, which, however, is not statistically significant. A
positive trend was also inferred for the adult population.
3. The attributive risk (a fraction of radiogenic cancers among all cancer
cases) derived from direct estimates of radiation risk for residents of the
Oryol oblast, is equal to 15% for children and adolescents at exposure and 4%
for adults. A conservative estimate of attributive risk of 60% has been adopted
by UNSCEAR and is used as a tentative estimate for children and adolescents of
the Oryol oblast, given the risk coefficient of 40 for the dose 1 Sv is used.
4. The performed analysis provides sufficient evidence to assume that the
radiation exposure factor did influence the thyroid cancer incidence in the
Oryol oblast, although to a lesser extent than in the
References
1.
Ivanov V.K., Gorski A.I., Tsyb A.F. et al. Dynamic of thyroid cancer incidence in
2.
Ivanov V.K., Gorski A.I., Pitkevitch V.A. et al. Risk of radiogenic thyroid cancer in
3.
Jacob J., Goulko G., Heidenreich W.F. et al. Thyroid cancer risk to children calculated//Nature. 
1998.  V. 392.  P. 3131.
4.
Jacob J., Kenigsberg, Zvonova I. et al. Childhood exposure due to the
5.
Likhtarev I.A, Sobolev B.G., Kairo I.A. et al. Thyroid cancer in the
6.
Likhtarev I.A, Kairo I.A., Shpak V.M. et al. Radiationinduced and background thyroid cancer of Ukranian children
(dosimetric approach)//Int. J. Radiat. Med.  1999.  V. 34.  P. 5166.
7.
Preston D.L., Lubin J.H., Pierce DA., McConney M.E. EPICURE. 
8.
Ivanov V.K., Gorski A.I., Tsyb A.F., Maksioutov M.A., Rastopchin E.M. Dynamics of thyroid cancer incidence in Russia following the Chernobyl
accident//J. Radiol. Prot.  1999.  V. 19, N 4.  P. 305318.
9.
Ivanov V.K., Gorski A.I., Maksioutov M.A. et al. Thyroid cancer incidence among adolescents and adults in the
10. Breslow N.E., Day N.E. Statistical methods in cancer