BioMed CentralEnvironmental Health

ss

Open AcceResearchExposure from the Chernobyl accident had adverse effects on erythrocytes, leukocytes, and, platelets in children in the Narodichesky region, Ukraine: A 6-year follow-up studyEugenia Stepanova1, Wilfried Karmaus*2, Marina Naboka3, Vitaliy Vdovenko1, Tim Mousseau4, Viacheslav M Shestopalov3, John Vena2, Erik Svendsen2, Dwight Underhill5 and Harris Pastides2

Address: 1Scientific Center for Radiation Medicine, Academy of Medical Sciences of Ukraine, Kyiv, Ukraine, 2Department of Epidemiology and Biostatistics, Norman J. Arnold School of Public Health, University of South Carolina, Columbia, South Carolina, USA, 3Radioecological Center, Ukrainian National Academy of Sciences, Kyiv, Ukraine, 4College of Arts and Sciences, University of South Carolina, Columbia, South Carolina, USA and 5Department of Environmental Health Science, Norman J. Arnold School of Public Health, University of South Carolina, Columbia, South Carolina, USA

Email: Eugenia Stepanova - profstepanova@i.ua; Wilfried Karmaus* - karmaus@sc.edu; Marina Naboka - marinavn1@yahoo.com; Vitaliy Vdovenko - xrisk06@yahoo.com; Tim Mousseau - mousseau@sc.edu; Viacheslav M Shestopalov - vsh@hydrosafe.kiev.ua; John Vena - Jvena@gwm.sc.edu; Erik Svendsen - svendsee@gwm.sc.edu; Dwight Underhill - dmunderhill@gmail.com; Harris Pastides - Pastides@gwm.sc.edu

* Corresponding author

AbstractBackground: After the Chernobyl nuclear accident on April 26, 1986, all children in the contaminatedterritory of the Narodichesky region, Zhitomir Oblast, Ukraine, were obliged to participate in a yearlymedical examination. We present the results from these examinations for the years 1993 to 1998. Sincethe hematopoietic system is an important target, we investigated the association between residential soildensity of 137Caesium (137Cs) and hemoglobin concentration, and erythrocyte, platelet, and leukocytecounts in 1,251 children, using 4,989 repeated measurements taken from 1993 to 1998.

Methods: Soil contamination measurements from 38 settlements were used as exposures. Blood countswere conducted using the same auto-analyzer in all investigations for all years. We used linear mixedmodels to compensate for the repeated measurements of each child over the six year period. Weestimated the adjusted means for all markers, controlling for potential confounders.

Results: Data show a statistically significant reduction in red and white blood cell counts, platelet countsand hemoglobin with increasing residential 137Cs soil contamination. Over the six-year observation period,hematologic markers did improve. In children with the higher exposure who were born before theaccident, this improvement was more pronounced for platelet counts, and less for red blood cells andhemoglobin. There was no exposure×time interaction for white blood cell counts and not in 702 childrenwho were born after the accident. The initial exposure gradient persisted in this sub-sample of children.

Conclusion: The study is the first longitudinal analysis from a large cohort of children after the Chernobylaccident. The findings suggest persistent adverse hematological effects associated with residential 137Csexposure.

Published: 30 May 2008

Environmental Health 2008, 7:21 doi:10.1186/1476-069X-7-21

Received: 23 November 2007Accepted: 30 May 2008

This article is available from: http://www.ehjournal.net/content/7/1/21

© 2008 Stepanova et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Page 1 of 13(page number not for citation purposes)

Environmental Health 2008, 7:21 http://www.ehjournal.net/content/7/1/21

BackgroundAn explosion at the Chernobyl nuclear power plant onApril 26, 1986, the worst accident in the history of nuclearpower, resulted in radioactive pollution of much of thesurrounding area. In the Ukraine, 2,293 villages andtowns with a population of 2.6 million inhabitants werecontaminated. A plume of radioactive fallout drifted overparts of Europe and reaching eastern North America, aswell. Ever since, the public in these areas has been exposedto radiation, both externally and internally via contami-nated locally-grown food, water and air.

Estimates of detrimental health effects from chronic radi-ation exposure vary widely [1]. Nearly 20 years after theChernobyl disaster the World Health Organization in areport of the UN Chernobyl Forum found no evidence foran increased incidence of leukemia [2]. However, thesame report found a complete lack of analytical studies inwhich dose and risks were estimated on an individuallevel. There were a few studies that analyzed white bloodcells but most were based on a small number of children,focused mainly on micronuclei, and were often inconclu-sive [3-8]. Lenskaia et al., analyzed blood smears from 820children living in a the Bryansk area in Russia and 46 con-trols from non-contaminated areas [9]. Using cytochemi-cal assays (mucopolysaccharids) and esterase in 464children with various exposure levels and 46 childrenfrom non-contaminated areas the work showed a reduc-tion of mature T-lymphocytes and an increase of imma-ture B-lymphocytes. In 1994–1996, Vykhovanets et al.and Chernyshov et al. studied T-lymphocytes in healthychildren, 219 and 120, respectively, and children sufferingfrom recurrent respiratory diseases (RRD) who residedaround Chernobyl. Both studies compared the exposedgroups with 148 non-exposed children, who were healthyor suffered from RRD. No information of leukocytecounts was provided [10,11].

Regarding red blood cells, we identified six publishedstudies [5,12-16]. Stepanova et al. found more transitory,prehemolytic and degenerative forms of erythrocytes (redblood cells) in exposed children in comparison with con-trol children [14]. Cross-sectional results on blood indicesfor years 1986, 1992 and 1998 were provided byBebeshko et al. [16]. The authors examined children in thefollowing age-groups: up to 3, 4–7 and 8–15 years old,residing in the Kiev, Zhytomyr, and Chernohiv provinceswith 137Cs soil contamination density of 37 kBq/m2 orless (37 kilo Bequerel/m2 = 1 Curie (Ci)/km2) and con-tamination densities between 38 and 55 kBq/m2. Theerythrocyte and leukocyte counts were significantlydecreased in children aged 0–3 years living in Zhytomyrand in children age 8–15 years living in Chernihiv. Theauthors found no differences in hemoglobin, erythrocyte,leukocyte and platelet count in children residing in settle-

ments with 137Cs soil contamination density of 38–555kBq/m2 compared to 37 kBq/m2.

The Chernobyl Sasakawa Health and Medical Coopera-tion project conducted the largest cross-sectional investi-gation of health effects in children [17]. Among a varietyof markers, hematological outcomes were determined in118,773 children residing in more than 80 regions withdifferent radiation exposures from 1991 to 1996, includ-ing 779 children from the Narodichi region [18,19]. Eachregion represented a distinct exposed population ratherthan a regional sample of the larger exposed population.Children from each region were examined at a differentpoints in time and there was no follow-up of eachregional cohort in subsequent years. Some children werere-examined, but only those who presented a critical clin-ical outcome. The Chernobyl Sasakawa Health and Medi-cal Cooperation project was a successful andcomprehensive screening project. However, the projectwas not sufficiently designed to assess differences inhealth effects in regions and over time. For instance, theresult of the exams in different regions and times, thoughpresented as time trends, were confounded by differentregions examined in different years [19]. In addition,improvements detected in the re-examinations of childrenwere judged as recovery. This ignores the fact that exclu-sive examination of observations above a critical valuewill by chance produce 'recovery' due to chance move-ments of observations below the cut-off point in the re-exam (regression to the mean). Although the results ofthis study have never been reported in peer-reviewed sci-entific journals, the project has been influential whenassessing the risk of the accident.

No study has yet investigated hematological follow-updata in children exposed from the Chernobyl accident.This motivated us to asses the association between resi-dential soil density of 137Caesium (137Cs) and hemo-globin concentration, and erythrocyte, platelet, andleukocyte counts in repeated measurements taken from1993 to 1998.

MethodsPopulationThe settlements in the Narodichesky region (ZhitomirOblast, Ukraine) are approximately 80 km from the Cher-nobyl nuclear site (Figure 1). Approximately 11,400 peo-ple reside in this region, including about 2,000 children[20]. Three quarters of its population live in rural villages,the others in small towns. The food supply is predomi-nantly locally grown. To monitor the health effects of theaccident, since 1986 every child from birth to age 18 in thecontaminated territory was required to have a yearly med-ical examination. The Human Subject Committee at theUniversity of South Carolina has approved these analyses.

Page 2 of 13(page number not for citation purposes)

Environmental Health 2008, 7:21 http://www.ehjournal.net/content/7/1/21

Page 3 of 13(page number not for citation purposes)

Map of the Narodychi region's territory soil pollution by 137CsFigure 1

Map of the Narodychi region's territory soil pollution by 137Cs.

Figure Legends

Environmental Health 2008, 7:21 http://www.ehjournal.net/content/7/1/21

Exposure measurementFollowing the Chernobyl accident, soil became contami-nated with 137Cs, a nuclear fission product that has a half-life of 30 years, and emits both beta- and gamma-irradia-tion (Figure 1). The highest levels of 137Cs were found inthe surface layers of the soil where it is absorbed by plantsand mushrooms, thereby entering the local food supply.Dosimetric assessments of settlements in the Ukraine con-taminated by radionuclides were published from 1991 to2005 in 10 reports [21]. In addition, the Ministry of Emer-gency Situations provided spatial data on radiation [22].Soil measurements used in this work represent the averageof numerous measurements in each village over severalyears, in particular 1991–1993. These averages were offi-cially named 'certifications' and documented in publica-tions by the Ministry of Health. [21]. Since residential andindividual 137Cs levels are highly correlated, we used thecontamination in the residential area as an approximationof individual exposure [23].

Medical examinationThe obligatory medical examination included weight andheight measurements, blood sampling, blood pressure,ultrasound measurements (thyroid, abdomen and kidneyand adrenals), lung function (starting with a height of 100cm), and a medical history. Gender and place of residencewere also noted. No other information on living condi-tions or tobacco smoke exposure was collected as a part ofthis exam. For these obligatory examinations, no parentalconsent was required.

This work is a first in a series of longitudinal analyses.Here we focus on hematologic function: erythrocytes (redblood cells), leukocytes (white blood cells), and throm-bocyte (platelet) count and hemoglobin concentrations.Blood was collected in tubes containing EDTA. A bloodcount, including erythrocytes, hemoglobin, leukocytes,and thrombocytes, was conducted using Sysmex model F-800 (TOA Medical Electronics Company, Kobe, Japan).Normal blood smears were stained by the standardizedazure B-eosin GIEMSA Y Romanowsky procedure. Results

of these investigations were entered into a database for theperiod of 1993 to1998.

Statistical AnalysisFor descriptive purposes, values beyond clinical referencelevels were defined in Table 1[24].

Data were analyzed using the statistical package SAS Ver-sion 9.1 (SAS Institute Inc., Cary, NC, USA). We had sixrepeated measurements of red and white blood cell mark-ers. To determine whether the measurements agreed overtime, we estimated intraclass correlation coefficients forthe repeated measurements [25].

To investigate the effect of residential 137Cs radiation, weused linear models for repeated measurements (PROCMIXED) [26]. This statistical method compensates for therepeated measurements of each child over the six cross-sectional models (1993–1998). For all markers, we esti-mated the adjusted means, controlling for confounders.

For the statistical estimation, the regular maximum likeli-hood method was applied. This model requires that therandom effects and the error vector are normally distrib-uted, which was found to be the case for all markers. Forthe within-subject association, modeling started with anunstructured covariance model which required least con-straints. For the repeated measurements, the initial modelused serial correlation structures (Gaussian). Based on theAkaike information criterion (AIC), we could simplify therandom effect to variance component, but could not sim-plify the repeated measurement matrix.

The residential measurements of 137Cs were ranked intofive groups of nearly equal size (PROC RANK), with eachrank treated as an indicator variable, facilitating assessingan exposure-response relationship. As for confoundingfactors, the statistical models included gender, age, andyear of measurement. Age was categorized into 18 indica-tor variables of one-year intervals, with the first group 0–1.5 years old. We used indicator variables for age and year

Table 1: Values beyond clinical reference levels.

Red blood cell count: younger than 2 years: ≥3.7 × 1012cells/L2 years and younger than 6: ≥3.9 × 1012cells/L6 years and younger than 12: ≥4.0 × 1012cells/L12 years and younger than 18: ≥4.5 × 1012cells/L

Hemoglobin: younger than 6 years: ≥5.59 mmol/L (9 g/dL)6 years and younger than 12: ≥7.14 mmol/L (11.5 g/dL12 years and older: ≥7.76 mmol/L (12.5 g/dL)

Thrombocytes: all ages: ≥150 × 109platelets/LLeukocytes: 1 to 3 years: ≥6–17.5 × 109cells/L

4 to 7 years: ≥5.5–15.5 × 109cells/L8 to 13 years: ≥4.5–13.5 × 109cells/L13 years and older: ≥4.5–11.0 × 109cells/L

Page 4 of 13(page number not for citation purposes)

Environmental Health 2008, 7:21 http://www.ehjournal.net/content/7/1/21

Table 2: Villages, soil contamination, number of participating children and measurements ψ

Village Quintile Soil contamination 137Caesium (kBq/m2) Children Repeated measurements

Number % (n = 1247) Number % (n = 4,981)

Brodnik 59 11 0.9 37 0.7

Slavkovci 67 9 0.7 36 0.7

Liplanchina 76

Vyasovka 79 60 4.8 262 5.3

Rubegivka 0 80 50 4.0 205 4.1

Radcha 82

Ovruch 102 4 0.3 14 0.3

N. Radcha 104

Klochki 105 25 2.0 79 1.6

Gresla 102 5 0.4 29 0.6

Norinci 112 71 5.7 305 6.1

Guto-Mariatyn 116 24 1.9 97 1.9

Babinichi 126 9 0.7 45 0.9

Laski 127 75 6.0 317 6.4

Jerev 1 130 28 2.2 111 2.2

Bolotnica 131 53 4.3 182 3.7

Davidki 132 8 0.6 38 0.8

Latashi 134 46 3.7 214 4.3

Snytiche 152 9 0.7 39 0.8

Bogdanovka 161

O. Dorogyn 165 51 4.1 220 4.4

Zakusily 165 35 2.8 120 2.4

N. Dorogin 169 22 1.8 71 1.4

Slavichina 170 7 0.6 30 0.6

Odruby 2 190 6 0.5 11 0.2

Zalesie 198 55 4.4 217 4.4

Suharevka 210 20 1.6 64 1.3

Page 5 of 13(page number not for citation purposes)

Environmental Health 2008, 7:21 http://www.ehjournal.net/content/7/1/21

of observation to investigate whether their relationshipswith the outcome variables were linear. If not, these indi-cator variables adjusted for non-linearity – for instance aswe did with erythrocyte counts and age in girls. To deter-mine whether the change in the distribution of the bloodmarkers over the time of the observation depends also onthe residential radiation level, we included an interactionterm of time (year of measurement) and 137Cs. In addi-tion, we investigated statistical differences in childrenborn before or after the accident.

ResultsChildren from 38 different settlements were included inthe cohort (Table 2). All villages had different soil meas-urements of 137Cs (Figure 1), ranging from 29 to 879 kBq/m2 (Figure 1). The average measurements shown for eachsettlement within colored zones are based on contamina-tion modeling. Average values and the colored zones wereretrieved from published data [21,22]. Three villages wereat the border or outside of the Narodichesky region(Bogdanovka, Gunichu, and Ovruch, total of 3 children).Due to the distribution of 137Cs, quintiles of the exposuredid not produce equally sized groups (Table 3).

The exact number of children residing in the Narod-ichesky region between 1993 to 1998 is unknown. Of the11,400 residents, approximately 2,000 children are chil-dren. Data of all exams (1993 to 1998) were included.Officially, participation in the exams was obligatory.However, authorities did not enforce participation.Hence, the term 'obligatory' primarily reflects the oppor-tunity of having annual examinations. Overall, approxi-mately 75% participated in at least one of the

examinations (1,459 of about 2,000 children residing inthe region). Of these, 1,247 children had a blood sampleanalyzed (86%), which resulted in 4,981 repeated meas-urements (Table 3). About one third of the children were4.5 years and younger, but only 15.9% of the measure-ments came from these children, indicating a lower partic-ipation in blood sampling. The oldest child was born in1979; the youngest in 1996. Data from the children bornin 1986 will be analyzed and reported in a separate paper.Of our cohort, 549 children were born before and 698after the accident (Table 3). In 75% of the available data,the first participation was documented in 1993, however,the measurements were nearly equally distributedthroughout the years. In 1993, we had 886 measurements;786 in 1994; 787 in 1995; 783 in 1996; 1,030 in 1997;and 709 in 1998.

Because erythrocyte count and hemoglobin concentrationmeasurement represent a nearly identical feature, thesemarkers are highly correlated (r = 0.72, Table 4). The othermarkers did not show high rank-correlations indicatingindependent measurements.

The intraclass-correlation coefficients (ICC) for the fourmarkers over the six years are: erythrocyte count: ICC = 0.4(5% confidence level: 0.37), for hemoglobin: ICC = 0.59(5% confidence level: 0.57), for thrombocyte count: ICC= 0.54 (5% confidence level: 0.52), and for leukocytecount: ICC = ICC = 0.53 (5% confidence level: 0.51).These results document substantial stability of the meas-urements.

Moteyki 224 29 2.3 132 2.7

Gunichu 225 12 1.0 32 0.6

Megeliska 236

Yajberen 253 19 1.5 90 1.8

Rudnya Basarskaya 3 266 116 9.3 486 9.8

Selec 310

Narodichi 350 279 22.4 1,049 21.1

Basar 4 364 103 8.3 433 8.7

Rossohivske 608 6 0.5 16 0.3

Hristinovka 879

ψ Villages with a small number of children/observations were merged for descriptive purposes.

Table 2: Villages, soil contamination, number of participating children and measurements ψ (Continued)

Page 6 of 13(page number not for citation purposes)

Environmental Health 2008, 7:21 http://www.ehjournal.net/content/7/1/21

Page 7 of 13(page number not for citation purposes)

Table 3: Characteristics of the study population

Children participating (n = 1,247) % Total number of observations (n = 4,981) %

Sexboys 49.1 48.4

Age groups (years) (first participation or year)up to 4.5 33.0 15.9>4.5 to 9.5 37.5 33.7>9.5 to 14.5 27.4 41.1>14.5 2.2 9.2

Year of birth1979 0.2 0.11980 4.6 4.21981 4.8 5.31982 8.3 9.71983 9.7 11.41984 9.3 12.01985 7.1 9.1#1987 6.7 8.11988 8.3 9.01989 8.5 7.71990 6.8 6.21991 6.3 5.31992 5.9 5.11993 5.0 3.31994 5.1 2.31995 2.7 1.01996 0.8 0.3

Year of first participation (or first entered into the data)1993 71.1 17.71994 2.4 15.81995 2.7 15.81996 1.6 15.71997 18.9 20.71998 3.4 14.3

Quintiles of the area contamination: 137Caesium (kBq/m2)29–112 18.9 19.4116–156 20.1 20.8165–253 9.3 19.4266–310 31.1 9.8350–879 18.9 30.7

# Children born in 1986 being analyzed and reported in a separate paper

Table 4: Median, minimal and maximal values, and rank correlation of hematological markers #

Variable Median Min Max Rank correlation (Spearman) and p-value

Hemoglobin Leukocyte Platelets

Erythrocyte count (1012L) 4 2.1 5.7 0.72 0.25 0.220.00 0.00 0.00

Hemoglobin (g/dL)Φ 12.3 5.2 16.9 0.18 0.160.00 0.00

Leukocyte count (106L) 6.8 2 18.9 0.350.00

Platelet count (109L) 252 108 670

# 4,981 repeated measurement of 1,247 childrenΦ 1 g/dL equals about 0.6206 mmol/L.

Environmental Health 2008, 7:21 http://www.ehjournal.net/content/7/1/21

Including all cohort children, the repeated measurementanalyses (mixed models) showed significant effects for theresidential 137Cs measurements (quintiles), the time ofthe medical examination, and the 137Cs × time interaction(Table 5). To illustrate the interactions of time and expo-sure, we extracted mean values for residential exposureclasses and years from the mixed models (see details inAdditional file 1).

The erythrocyte count shows a minor increase from 1993to 1998 (Figure 2), starting from below 4 × 1012 cells perliter to more than 4.1 × 1012 cells per liter. The improve-ment is less in the two groups of children with higher res-idential exposures; thus the exposure gradient becomesstronger in 1998 compared to 1993. A similar trend isobvious for hemoglobin. Both markers show a dip in1996, which is not explained by age in general nor by aparticular age in girls (age at menarche). For plateletcounts, it is obvious that children in all 137Cs exposuregroups improved over time (Figure 3). The exposure gra-dient is stronger in 1993, but diminished in 1998. Thereseems to be no further increase after 1997. Figures 2 and3 show lower values in 1996 in all exposure groups.

There appears to be exposure-response patterns in allgraphs; however, the second to the highest exposuregroup showed the strongest effect in all four outcomes.

Compared to the second highest group, the highest resi-dential exposure group includes a higher proportion ofsemi-rural children, from Narodichi, the administrativecenter of the region (Table 2).

For leukocyte counts, the exposure gradient did notchange between 1993 and 1998 (Table 6). Starting with asoil contamination of about 165 kBq/m2, the leukocytecount is approximately 1 million cells per liter lower inthe group of children with the highest soil contamination.A similar reduction was found for children born after theaccident in 1986 (data not shown). However, the countincreased over the time period from 1993 to 1996.

Children born after the accident and thus exposed duringpregnancy and childhood showed no 137Cs × time interac-tion (Table 7). For the three outcomes, erythrocyte andplatelet count and hemoglobin concentration, the valuesare significantly lower in the two highest residential expo-sure groups; the effect is strongest for platelets and whiteblood cells. Compared with the measurements in 1998there is an overall improvement over the six years ofobservation for erythrocyte, platelet, but no clear trendregarding hemoglobin (data not shown).

Fifty-three percent of all erythrocyte counts were belowthe clinical age-specific reference values, as were 19% of

Table 5: Statistics of the models including the interaction of area contamination and year of measurement #

Numerator degree of freedom Denominator degree of freedom F-Value

Prob F

Erythrocyte countQuintiles of the area contamination with Caesium 137 (kBq/m2)

4 3692 12.55 <0.001

Sex 1 3692 12.11 <0.001

Age groups 17 3692 13.42 <0.001

Year of measurement 5 3692 102.42 <0.001

Quintiles of the area contamination: 137Caesium (kBq/m2) × year

20 3692 5.37 <0.001

HemoglobinQuintiles of the area contamination with Caesium 137 (kBq/m2)

4 3692 7.28 <0.001

Sex 1 3692 7.72 0.006Age groups 17 3692 33.43 <0.00

1Year of measurement 5 3692 52.68 <0.00

1Quintiles of the area contamination: 137Caesium (kBq/m2) × year

20 3692 4.81 <0.001

Platelet countQuintiles of the area contamination with Caesium 137 (kBq/m2)

4 3690 9.71 <0.001

Sex 1 3690 1.61 0.20

Page 8 of 13(page number not for citation purposes)

Environmental Health 2008, 7:21 http://www.ehjournal.net/content/7/1/21

the hemoglobin measurements, 0.7% of the platelet and26.6% of the white blood cell counts. The proportion ofvalues below the clinical limit decreased from 1993 to1996. The white blood cell counts also showed improve-ment, but an increasing proportion of mild leukocytosis(white blood cell count > 11 × 109cells per L; 1993: 0.1%,1997: 3.2%, 1998: 2.0%).

DiscussionWe analyzed markers of hematopoiesis from 1993 to1998 and their association with residential radiationexposure and found an adverse effect on erythrocyte,platelet, and white blood cell counts and on hemoglobinconcentration. When analyzing the total sample of chil-dren irrespective of whether the child was born before orafter the accident (exposure occurred in childhood vs. inutero and in childhood), we found that the recovery ofplatelets was more pronounced in children who wereexposed to higher residential contamination with 137Cs.The recovery of erythrocytes and hemoglobin was smallerin the highly exposed children. White blood cell counts

Changes of erythrocyte counts and hemoglobin over the years of observations by 137Cs exposureFigure 2Changes of erythrocyte counts and hemoglobin over the years of observations by 137Cs exposure. The 137Cs contamination is grouped into quintiles. For hemoglobin: 1 g/dL equals about 0.6206 mmol/L.

3.8

3.85

3.9

3.95

4

4.05

4.1

4.15

4.2

4.25

1993 1994 1995 1996 1997 1998

Red blood cell count (1012/L) 1,247 children, 4,981 observations

29-112 137Cs (kBq/m2)

116-156

165-253

266-310 137Cs (kBq/m2)

350-879

11.8

12

12.2

12.4

12.6

12.8

1993 1994 1995 1996 1997 1998

Hemoglobin (g/dL) 1,247 children, 4,981 observations

29-112 137Cs (kBq/m2)

116-134

165-253

266-310 137Cs (kBq/m2)

350-879

Changes of platelet counts over the years of observations by 137Cs exposureFigure 3Changes of platelet counts over the years of observa-tions by 137Cs exposure. The 137Cs contamination is grouped into quintiles.

200

210

220

230

240

250

260

270

280

290

300

1993 1994 1995 1996 1997 1998

29-112 137Cs (kBq/m2)

116-156

165-253

350-879 266-310 137Cs (kBq/m2)

Platelet count (109/L) 1,247 children, 4,981 observations

Page 9 of 13(page number not for citation purposes)

Environmental Health 2008, 7:21 http://www.ehjournal.net/content/7/1/21

did not show an exposure × time interaction, butremained lower in higher exposed children. Also in chil-dren born after the accident there was no such interactioneffect. An outcome improvement was obvious in all expo-sure groups, with the exception of hemoglobin, but theexposure gradient did not diminish over the observationperiod.

Overall, approximately 75% participated in the examina-tions (1,459 of about 2,000 children residing in theregion); blood samples were available for 63%. Officially,participation in the exams was obligatory. However,authorities did not enforce participation.

This is a dynamic cohort study: children were entering andleaving the study. Regarding non-participation in oneexam, lower or higher levels of the three blood count var-iables in the preceding exam were not associated withmissing the next exam. Hence, we have no indication fora selection bias due to differential attrition.

The justification to enter and analyze only data for 1993to 1998 was due to budgetary limitations. We are seekingfunding to extend the data entry and include morerepeated measurements. The 1993 to 1998 period waschosen since all exams were established during this time-frame. Another advantage of this period is that it includesa comparable number of children who were born beforethe accident in 1986 and thereafter. This facilitates a com-parison of exposures occurring after birth in children bornbefore the accident in 1986 with exposure in childrenborn after the accident (persistent 137Cs exposure). Thevariable "born before" addresses two aspects, first,

whether the child was exposed to other short-lived radio-nuclides immediately after the accident (for instance131Iodine, in children born before 1986), and, second,whether the child was exposed to 137Cs during pregnancyand in its development (born after).

Originally, we planned to analyze radiation exposureusing individual effective equivalent dose in millisievert(mSv), which incorporates total external and internaldoses, including direct measurements with whole bodycounts [27], but there are uncertainties in the time periodsin which the measurements were taken and their docu-mentation. In addition, the individual effective equivalentdose was not calculated for every child and exposure mod-els for individual dosimetry were changed over time.Although residential radiation exposure introduces morenon-differential misclassification and thus weakens possi-ble associations, those data were our most complete avail-able exposure data set. In addition, the ChernobylSasakawa Health and Medical Cooperation project foundthat individual 137Cs levels in the bodies of children werehighly correlated (r = 0.7, p < 0.01) with the contamina-tion level in the place of residence [23]. However, com-pared to the five groups of residential exposure, theindividual effective equivalent dose is highest in childrenin the second to highest residential exposure group (meanindividual effective equivalent doses over increasing levelsof residential exposure groups: 13.6, 11.7, 16.7, 34.9, 19.6mSv). Thus, it is likely that the stronger effect in the sec-ond highest residential exposure (266–310 137Cs kBq/m2)follows from the higher individual dose in this group. Wenote that the highest exposure group represents childrenfrom Narodichi, the largest village of the region (Table 2),

Table 6: Estimate effects of 137Caesium and year of measurements on white blood cell counts Ψ

White blood cell count (106/L) 1,247 children, 4,981 measurement

Effect Adj. Mean# Lower Upper p

Quintiles of the area contamination: 137Caesium (kBq/m2) 29–112 6.87 6.58 7.16 Ref.116–156 6.88 6.60 7.18 0.94165–253 6.40 6.11 6.68 0.01266–310 5.95 5.55 6.35 <0.001350–879 5.81 5.57 6.04 <0.001

Ψ P-values are based on t-test compared with the reference (Ref.).# Adjusted for age, sex, born after the accident, and year of measurement

Age groups 17 3690 5.58 <0.001

Year of measurement 5 3690 92.61 <0.001

Quintiles of the area contamination: 137Caesium (kBq/m2) × year

20 3690 2.08 0.003

# For the illustration of the results see Figure 2 and 3.

Table 5: Statistics of the models including the interaction of area contamination and year of measurement # (Continued)

Page 10 of 13(page number not for citation purposes)

Environmental Health 2008, 7:21 http://www.ehjournal.net/content/7/1/21

whereas the second to highest exposure group includesmore children from small villages and rural areas with alarger chance of direct exposure externally and internallyfrom food. Furthermore, compared to the virtue of indi-vidual measurements, using residential exposure data hasthe advantage that it provides the information needed tocontribute to improved exposure regulations.

There are other sources of non-differential exposure mis-classification, e.g., in landscapes with relatively low levelsof radioactive contamination of the soil, the populationmay receive substantial radiation doses due to their con-sumption of contaminated food and vice versa. Theresults also show decreased erythrocyte counts and hemo-globin concentrations in 1996 in all exposure groups (Fig-ures 2 and 3). It is possible that there was an instrumentbias in 1996. Another explanation is that the winter in1996 was colder. Thus, children may have suffered from areduced supply of vitamins that are essential for thehematopoesis. Since the decline occurred in all exposuregroups, this represents a non-differential misclassifica-tion. Non-differential misclassifications tend to produceresults that underestimate effects and do not present athreat to validity of the findings.

Our data provide only a little information on individualcharacteristics such as age and gender. One may beinclined to consider that other confounding factors suchas parental smoking or diet would increase the internalvalidity of this study. Against that, we need to understandthe setting of this research. The Chernobyl accidentaffected all groups of the population, irrespective ofwhether the parent exposed their children to second handsmoke or whether their diet was healthy or not. Thus,smoking and other factors cannot act as confounders,since the accident resulted in a random distribution ofradiation in relation to various other health related risk

factors. Such a setting is referred to as quasi-experimentalor as a randomized natural experiment [28,29]. Althoughthis setting does not require the control of confounders,we cannot exclude other risk factors that may have mod-erated the adverse effects of radiation (effect modificationor interaction). For instance, it is possible that poverty andunhealthy lifestyles have augmented adverse radiationeffects. The sample of children was homogeneous. Allfamilies were poor villagers with a traditional diet basedon local food and comparable domestic conditions.Given the setting of a randomized natural experiment, ourresults provide compelling evidence for the adverse effectsof residential 137Cs radiation on hematopoiesis.

Another limitation is that we have only computerized andstatistically analyzed data for a limited time-period(1993–1998). However, this is a first step, and there is aclear need to include additional data before 1993 andafter 1998. In addition, we had to start with a specific setof health outcomes and chose markers related to hemat-opoiesis. Other reports that document the various aspectsof the Chernobyl accident will follow. These childrenseem to suffer from multiple diseases and co-morbiditieswith repeated manifestations, a condition for which theterm 'frequently sick child' was defined by Stepanova andcoworkers [30].

According to data from the Ukrainian Ministry of SocialProtection (January 2005) and Ministry of Emergency,more than half a million children reside in areas withchronic exposure to low radiation due to soil contamina-tion with 137Cs. We believe that it is extremely importantto analyze not only cancer-related outcomes but also non-neoplastic effects [31], which are much more frequentthan cancer. It is surprising that the UN Chernobyl ForumReport [2] did not consider multiple publications byUkrainian, Russian, and Byelorussian researchers about

Table 7: Estimated effects of 137Caesium and year of measurements on erythrocyte count, hemoglobin, and platelet counts Ψ

Erythrocyte count (1012/L) 698 children, 2,404 measurements

Hemoglobin (g/dL)Φ 698 children, 2,404 measurements

Platelet count (109/L) 698 children, 2,403 measurements

Effect Adj. Mean# 5–95% confidence

interval

p Adj. Mean# 5–95% confidence

interval

p Adj. Mean# 5–95% confidence interval

p

Quintiles of the area contamination: 137Cs (kBq/m2)

29–112 4.02 3.98- 4.06 Ref. 12.2 12.1- 12.3 Ref. 284 274 294 Ref.

116–156 3.99 3.97- 4.03 0.29 12.0 11.9- 12.2 0.039 273 266 282 0.07165–253 4.00 3.93- 3.93 0.42 12.1 12.0- 12.2 0.34 281 273 290 0.52266–310 3.88 3.84- 3.93 <0.001 11.8 11.7- 12.0 0.001 255 242 267 <0.001350–879 3.96 3.93- 3.99 0.009 12.0 11.0- 12.1 0.033 264 256 271 <0.001

Φ 1 g/dL equals about 0.6206 mmol/L.Ψ P-value are based on t-test compared with the reference (Ref.) in children born after the accident.# Adjusted for age, sex, and year of measurement

Page 11 of 13(page number not for citation purposes)

Environmental Health 2008, 7:21 http://www.ehjournal.net/content/7/1/21

the excess of non-cancer morbidity in children living inthe territory contaminated by the Chernobyl accident[9,32-37]. Because the Chernobyl accident was like fewothers, there is a gap in knowledge regarding potentialhealth sequelae. The atomic bombing of Japan near theend of World War II did release fission isotopes over awide area, but the composition of this isotopic contami-nation was different [38]. The Chernobyl accidentinvolved the release of isotopes built up in the fuel rodsover time; these isotopes are far more persistent. As apotential consequence, the decreased blood counts showlasting effects. This was not observed after non-persistentradiation exposure [39]. However, similar effects werereported for the River Techa accident, which happened in1957 [40]. The exposure was characterized by gamma andbeta irradiation due to 90Strontium and 137Cs. As in chil-dren residing in Narodichesky region, adverse hemato-logic effects were detected. A normalization of thehematologic outcomes occurred only 13 years after theaccident [40]. In investigating health sequelae in a cohortof children exposed to the Chernobyl accident, we expectto find associations that have not been reported before.This work is the first in a series using longitudinal epide-miology to uncover long-term effects of the Chernobylaccident on children, leading to a better understanding ofhow large and persistent the radiation exposure affects thegeneral population.

ConclusionMore than 10 years after the Chernobyl accident, childrenin the Narodichesky region, Ukraine, approximately 80km from Chernobyl, showed decreased counts for red andwhite blood cells and platelets, and a reduced concentra-tion of hemoglobin associated with persistent residential137Cs exposure. There are compelling reasons to investi-gate more closely the relationship of radioactive exposureafter the Chernobyl accident and health sequelae in chil-dren.

AbbreviationsIntraclass correlation coefficient (ICC), ethylene diaminetetraacetic acid (EDTA)

Competing interestsThe authors declare that they have no competing interests.

Authors' contributionsEugenia Stepanova and Vitaliy Vdovenko have contrib-uted to developing the protocol of the medical exams andconducted the examinations and blood analyses. WilfriedKarmaus, Marina Naboka, Vitaliy Vdovenko, Tim Mous-seau, John Vena, and Harris Pastides developed the ana-lytical plan. Marina Naboka, Viacheslav M. Shestopalov,and Dwight Underhill provided access and supported theassessment of exposure data. Wilfried Karmaus and Erik

Svendsen analyzed the data. All the authors contributed toand approved the final manuscript.

Additional material

AcknowledgementsFunding for the data entry was provided by the Samuel Freeman Charitable Foundation, the Walker Institute for International and Area Studies, and the National Science Foundation. We thank Susan Davis, Curtis Travis, Dmitry Afanasyev for their comments and and Wanzer Drane for the initial steps of the data collection.

References1. Kinley D: Chernobyl's Legacy: Health, Environmental and

Socio-Economic Impacts and Recommendations to the Gov-ernments of Belarus, the Russian Federation and Ukraine.Second revised version. 2006 [http://www.iaea.org/Publications/Booklets/Chernobyl/chernobyl.pdf].

2. Bennett B, Repacholi M, Carr Z: Health Effect of the ChernobylAccident and Special Health Care Programmes. Report ofthe UN Chernobyl Forum Expert Group "Health". 2006[http://www.who.int/ionizing_radiation/chernobyl/WHO%20Report%20on%20Chernobyl%20Health%20Effects%20July%2006.pdf].

3. Vasil'eva IM, Unzhakov SV, Meliksetova IA, Zasukhina GD, Semia-chkina AN, Kazantseva LZ: [The adaptive response in the lym-phocytes of children from an area with an elevatedbackground of ionizing radiation]. Radiats Biol Radioecol 1995,35:662-665.

4. Tedeschi B, Caporossi D, Vernole P, Padovani L, Mauro F: Dohuman lymphocytes exposed to the fallout of the Chernobylaccident exhibit an adaptive response? III. Challenge withbleomycin in lymphocytes from children hit by the initialacute dose of ionizing radiation. Mutat Res 1996, 354:77-80.

5. Fenech M, Perepetskaya G, Mikhalevich L: A more comprehensiveapplication of the micronucleus technique for biomonitoringof genetic damage rates in human populations – experiencesfrom the Chernobyl catastrophe. Environ Mol Mutagen 1997,30:112-118.

6. Makedonov GP, Tskhovrebova LV, Unzhakov SV, Semiachkina AN,Vasil'eva IM, Zasukhina GD: [Radioadaptive response in lym-phocytes of children living in territories polluted by radionu-clides as a result of the accident at the Chernobyl powerplant]. Radiats Biol Radioecol 1997, 37:640-644.

7. Padovani L, Stronati L, Mauro F, Testa A, Appolloni M, Anzidei P,Caporossi D, Tedeschi B, Vernole P: Cytogenetic effects in lym-phocytes from children exposed to radiation fall-out afterthe Chernobyl accident. Mutat Res 1997, 395:249-254.

8. Mikhalevich LS, De Zwart FA, Perepetskaya GA, Chebotareva NV,Mikhalevich EA, Tates AD: Radiation effects in lymphocytes ofchildren living in a Chernobyl contaminated region of Bela-rus. Int J Radiat Biol 2000, 76:1377-1385.

9. Lenskaia RV, Buiankin VM, Zakharova GA, Ikonnikova OA, KoliaskinaMG, Popova OA, Rumiantsev AG, Seleznev VA, Sterligova IA, Cher-nov VM: [Cytochemical lymphocytic markers as criteria suit-able for mass and selective hematologic screening ofchildren living in areas contaminated by radionuclides].Gematol Transfuziol 1993, 38:34-36.

Additional file 1Table with confidence limits. Estimated combined effects of 137Caesium irradiation and year of measurements on erythrocyte count, hemoglobin, and platelet counts and their confidence limits.Click here for file[http://www.biomedcentral.com/content/supplementary/1476-069X-7-21-S1.doc]

Page 12 of 13(page number not for citation purposes)

Environmental Health 2008, 7:21 http://www.ehjournal.net/content/7/1/21

Publish with BioMed Central and every scientist can read your work free of charge

"BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime."

Sir Paul Nurse, Cancer Research UK

Your research papers will be:

available free of charge to the entire biomedical community

peer reviewed and published immediately upon acceptance

cited in PubMed and archived on PubMed Central

yours — you keep the copyright

Submit your manuscript here:http://www.biomedcentral.com/info/publishing_adv.asp

BioMedcentral

10. Vykhovanets EV, Chernyshov VP, Slukvin II, Antipkin YG, Vasyuk A,Colos V: Analysis of blood lymphocyte subsets in children liv-ing around Chernobyl exposed long-term to low doses ofcesium-137 and various doses of iodine-131. Radiat Res 2000,153:760-772.

11. Chernyshov VP, Vykhovanets EV, Slukvin II, Antipkin YG, Vasyuk AN,Strauss KW: Analysis of blood lymphocyte subsets in childrenliving on territory that received high amounts of fallout fromChernobyl accident. Clin Immunol Immunopathol 1997, 84:122-128.

12. Livingston GK, Jensen RH, Silberstein EB, Hinnefeld JD, Pratt G, Big-bee WL, Langlois RG, Grant SG, Shukla R: Radiobiological evalua-tion of immigrants from the vicinity of Chernobyl. Int J RadiatBiol 1997, 72:703-713.

13. Vartanian LS, Gurevich S, Kozachenko AI, Nagler LG, Burlakova EB:[Age-related peculiarities of effect of low dose ionizing radi-ation on blood antioxidant enzyme system status in Cherno-byl's accident liquidation participant]. Adv Gerontol 2004,14:48-54.

14. Stepanova EI, Davidenko OA, Vdovenko V: [Peculiarities of super-ficial architectonics of peripheral blood erythrocytes in chil-dren irradiated in utero and living in the radioactive-contaminated areas]. Tsitol Genet 2006, 40:40-43.

15. Spirichev VB, Donchenko GV, Blazheevich NV, Parkhomenko Iu M,Aleinik SI, Golubkina NA, Vrzhesinskaia OA, Isaeva VA, KodentsovaVM, Pereverzeva OG, Sokol'nikov AE, Iakushina LM: [To the 20thanniversary of the Chornobyl accident study of vitamin sta-tus and provision with micro- and macroelements of limitedgroups of people at different time periods since the accidentat Chornobyl nuclear power plant]. Ukr Biokhim Zh 2006,78:5-26.

16. Bebeshko VG, Bruslova EM, Tsvetkova NM: Hemopoietic systemstate in children under the chronic radiation impact withinlow dose range. In Clinical effects of the Chernobyl disaster Edited by:Bebesko VG, Kovalenko AN. Kiev: Academy of Medical Science;1999:21.

17. Yamashita S, Shibata Y, (Eds): Chernobyl: A Decade. Proceedings of theFifth Chernobyl Sasakava Medical Cooperation Symposium, 14–15 October1996 Kiev, Ukraine: Elsevier; 1996.

18. Avramenko A: Health conditions of children in the Kiev region10 years after the Chernobyl nuclear power plant accident:based on the results of health examination made by theChernoby Sasakawa Diagnostic Center (Kiev Regional Hos-pital No. 2). In Chernobyl: A Decade Proceedings of the Fifth ChernobylSasakava Medical Cooperation Symposium, Kiev, Ukraine, 14–15 October1996 Edited by: Yamashita S, Shibata Y. Amsterdam, Lausanne, NewYork: Elsevier; 1997:3-10.

19. Karevskaya IV, Fokina MM, Kozyreva EA, Drozdovich VS, AverichevAA, Demidenko AN, Derzhtskaya EV, Cot VA, Moiseyenko NV, Kou-likova NV, Kroupnik TA, Sribnaya VD, Tkachuk LP, Avramenko AI,Petrova AA, Voropai LV, Masyakin VB, Daniliuk LV, Daniliuk VV:Hematological finding of the Chernoby Sasakawa Health andMedical Cooperation Project. In Chernobyl: A Decade Proceedingsof the Fifth Chernobyl Sasakava Medical Cooperation Symposium, Kiev,Ukraine, 14–15 October 1996 Edited by: Yamashita S, Shibata Y.Amsterdam, Lausanne, New York: Elsevier; 1997:45-57.

20. All-Ukrainian population census '2001 [http://www.ukrcensus.gov.ua./eng/results/general/urban-rural/Zhytomyr]

21. Ministry of Health of the Ukraine: Certification doses for the settlementsof Ukraine located in contaminated territories after the Chernobyl accident,Reports No 1 to 7 and 9 and 10. Kiev 2005.

22. Ministry of Emergency Situation of Ukraine: Atlas of radioactivity con-tamination of the Ukraine. Kiev 2002.

23. Sharifov VF, Koulikova NV, Voropai LV, Kroupnik TA, Masyakin VB,Cot VA, Kovalev AI, Aksenov AS, Averichev AA, Moiseyenka MY,Avramenko AI, Goncharenko OE, Daniliuk VV: Finding of theChernoby Sasakawa Health and Medical CooperationProject: 137 Cs concentrations among children aroundChernobyl. In Chernobyl: A Decade Proceedings of the Fifth ChernobylSasakava Medical Cooperation Symposium, Kiev, Ukraine, 14–15 October1996 Edited by: Yamashita S, Shibata Y. Amsterdam, Lausanne, NewYork: Elsevier; 1997:39-44.

24. Nicholson JF, Pesce M: Reference Ranges for Laboratory Testsand Procedures. In Nelson textbook of pediatrics 16th edition. Editedby: Behrman RE, Kliegman RM, Jenson HB. Philadelphia: W.B. Saun-ders Company; 2000:2181-2229.

25. Armstrong B, White E, Saracci R: Principles of Exposure Measurementin Epidemiology Oxford, New York, Tokyo: Oxford University Press;1992.

26. Verbeke G, Molenberghs G: Linear Mixed Models for Longitudinal DataNew York: Springer-Verlag; 2000.

27. International Programme on the Health Effects of the ChernobylAccident: Health Consequences of the Chernobyl Accident. Results of theIPHECA Pilot Projects and Related National Programmes. Scientific Report.Geneva 1996.

28. Goldberg J, Richards MS, Anderson RJ, Rodin MB: Alcohol con-sumption in men exposed to the military draft lottery: a nat-ural experiment. J Subst Abuse 1991, 3:307-313.

29. Carrozzi L, Giuliano G, Viegi G, Paoletti P, Di Pede F, Mammini U,Carmignani G, Saracci R, Giuntini C, Lebowitz MD: The Po RiverDelta epidemiological study of obstructive lung disease:sampling methods, environmental and population charac-teristics. Eur J Epidemiol 1990, 6:191-200.

30. Stepanova EI, Kolpakov I, Vdovenko VY: State of the functional systemin children exposed to radiation effect as a result of the Chernobyl catastro-phe Kiev: Chernobyl Inform Press; 2003.

31. Busby C, Fucic A: Ionizing radiation and children's health: con-clusions. Acta Paediatr Suppl 2006, 95:81-85.

32. Astakhova LN, Demidchuk EP, Davydova EV, Arinchin AN, Gres NA,Zelenko SM, Drozd VM, Poliakova TI, Dardynskaia IV, Bazyl'chik SV:[Health status of Byelorussian children and adolescentsexposed to radiation as consequence of the Chernobyl AESaccident]. Vestn Akad Med Nauk SSSR 1991, 11:25-27.

33. Stepanova EI, Kondrashova VG, Davidenko OA, Kniazeva OB, Gali-chanskaia T: [The hematological effects of the Chernobyl acci-dent in children]. Gematol Transfuziol 1992, 37:31-33.

34. Stepanova EI, Chaialo PP, Kolpakov IE, Kondrashova VG, Kurilo LV,Galichanskaia T, Fediakova OB, Ivaniuk AG: [Effects of the conse-quences of the Chernobyl accident on the child's body]. Pedi-atria 1991, 12:8-13.

35. Lukyanova EM: Klinical and practical aspects of pediatrichealth nine years after the Chernobyl disaster. JAMS of Ukraine1996, 1:21-23.

36. Bobylova OA: Health state estimate in children depending onthe occured radio-ecological situation after the ChNPP acci-dent. Ukraine Radiology Journal 1996, 1:21-23.

37. Nyagu AI, Loganovsky KN, Nechaev SY, Pott-Born R, Repin VS,Antipchuk EY, Bomko MA, Yurryev KL, Petrova IV: Effects of pre-natal brain irradiation as a result of Chernobyl accident. Int JRadial Med 2004, 6:1-4.

38. Baverstock K, Williams D: The chernobyl accident 20 years on:an assessment of the health consequences and the interna-tional response. Environ Health Perspect 2006, 114:1312-1317.

39. Scherthan H, Abend M, Muller K, Beinke C, Braselmann H, Zit-zelsberger H, Kohn FM, Pillekamp H, Schiener R, Das O, Peter RU,Herzog G, Tzschach A, Dorr HD, Fliedner TM, Meineke V: Radia-tion-induced late effects in two affected individuals of theLilo radiation accident. Radiat Res 2007, 167:615-623.

40. Akleyev AV, Kossenko MM, Silkina LA, Degteva MO, YachmenyovVA, Awa A, Akiyama M, Veremeyeva GA, Vozilova AV, Kyozumi S:Health effects of radiation incidents in the southern Urals.Stem Cells 1995, 13(Suppl 1):58-68.

Page 13 of 13(page number not for citation purposes)