Introduction

Childhood cancers

Childhood Cancers (CC) are found among all cancers at a ratio of 2 %. 10 % of deaths during whole childhood period below 15 years old are due to CC [17]. Eight thousand children get cancer diagnosis each year in USA. Leukemia and brain tumors are the most prevalent cancers (30 ve 20 %, respectively). In recent years, deaths due to childhood cancer are decreasing but there is a trend for increase in cancer incidence. Incidence of ALL (Acute Lymphoblastic Leukemia) has increased by 27.4 % from 1973 until 1990.

Childhood cancers have many differences from adulthood cancers in clinical, biological and genetic aspects. In a study performed in USA, the incidence of a new cancer in a population including one million people each year was reported to be 129.77 children. It occurs between 0 and 5 years old more often. It was found that the incidence of CC varies depending on the countries and geographical regions. For instance; ALL is seen more in China, Japan and USA, whereas it occurs less in Middle East and Africa. Again, lymphoma are more frequently seen in Africa, but it occurs less in Japan [17].

The causes of childhood cancers are still being studied by many scientists [14]. The most significant factors that are emphasized on this topic are listed below.

  1. 1.

    Familial inheritance

  2. 2.

    Ultraviolet radiation

  3. 3.

    Ionizing radiation

  4. 4.

    Electromagnetic waves

  5. 5.

    Chemical agents

  6. 6.

    Viral infections

  7. 7.

    Diet

  8. 8.

    Interactions between children and environmental materials [2, 4, 7]. Early diagnosis has a significant importance in CC. Families have important responsibilities in this aspect [9].

Geographical information systems and their use in epidemiology

Geographical Information Systems (GIS) have a very wide application area. GIS is an information system carring out the functions as collection, storage, processing and presentation of geographical and attribute information obtained by geospatial observations [24]. Although development of GIS is based on 1960s in modern means, the first example of GIS application is the epidemiological application created by John Snow in order to detect cholera epidemic in London (UK). Snow marked cholera cases emerged in 1854 on the map by lines for each house. He thought that cholera bacteria caused these cases and determined that the nearest water pump near these cases was a pump in Broad Street and these people provided their drinking water from this pump. He applied to municipality to ask for blocking this pump. After the blockage, a significant decrease was seen in cholera cases [19]. Snow’s study is a simple but effective example that reveals the importance of map and thus, the use of geospatial information in epidemiological studies.

Geospatial information for different classes (e.g. buildings, roads, satellite images, etc) is expressed by separate layers in GIS (Fig. 1). GIS is a highly useful and necessary tool that performs procedures such as combining different numerical data, making geospatial queries, being suitable for automation, visual presentation, manipulation, making geostatistical analyses, giving prospective decisions and estimating possible situations against designed scenarios [11].

Fig. 1
figure 1

Examples of data layers in GIS

Full size image

Although the importance of GIS is increasing in epidemiological studies day by day, use of GIS in epidemiological studies is not still carried out systematically and within a particular policy framework. For instance; when cancer-related studies are examined, it is observed that these are some national projects carried out by universities and they do not show continuity. Examples from these studies on cancer are as follows: in a study by Colak and Yomralioglu [8] and Yomralıoğlu et al. [25] performed in Trabzon city, a possible relationship between cancer types and land cover (such as agriculture, forest, residential) and heights was investigated. The distribution of cancer cases in Konya city in 2006 was examined by GIS by Özkan and Güngör [16]. Güngör and Özkan [13] have performed a study on cancer cases in Kutahya in 2007 and on the distribution of death rates. Türk [21] strives to investigate cancer cases in Sivas between 2007 and 2010.

Purpose of the study

This paper presents the first results of an epidemiological study which has continuity. Zonguldak has hard coal mining area that do not only provide energy requirement for the iron and steel plants in Kdz. Eregli and Karabuk and two thermal power plants in Catalagzi and new stations planned but also for whole country. Use of this high calorific fuel for heating houses and lack of taking many environmental precautions cause significant contamination in air, soil and water (sea and drinking water) in overall Zonguldak. Kdz. Eregli has one of the largest iron-steel industry of Turkey, and Caycuma is an important area with small and medium sized industrial enterprises for the province. As understood from the studies about CC, environmental and chemical factors are important factors in the development of CC. It is required to give attention to the regions including the cases and the environmental characteristics of that region. Therefore, the aim of this study is to examine cancer cases observed in children living in Zonguldak epidemiologically with the help of GIS.

Materials and methods

Type of the study

This is a descriptive epidemiological study evaluating cancer cases with GIS.

Universe of the study and sample selection

Universe of the study is composed of 60 children between 1 and 19 years old who were diagnosed as cancer and treated in Children Oncology Clinic in Bulent Ecevit University between 15.03.2014 and 31.01.2016. Whole universe was reached in the study without selecting a sample.

Data collection tools

A form which was prepared by taking expert advice and which is composed of a total of 20 items and including sociodemographic characteristics, existing disease and treatment-related conditions of the children and their families was used as the data collection tool. Form was applied to the children and the parents in children oncology clinic at the dates of the study.

Analysis of data

SPSS 16 for Windows was used for data analysis. Results were expressed as percentages. Chi-Square test was used for intergroup comparisons. Results were assessed within 95 % confidence interval and p < 0.05 was considered as statistically significant.

Database of the study and the map showing counties of Zonguldak province were associated in ArcGIS software, and distribution of cancer cases were presented in map format. Variables used in the study were evaluated, recorded in prepared data collection form and distribution maps were produced.

Ethical aspect of the study

All necessary institutional permits and ethics committee approval were taken in order to perform the study.

Results

It was found that 35 % (n = 21) of the children included in the study were between 5 and 8 years old, 51.7 % (n = 31) were males, 48.3 % (n = 29) were females, 36.7 % (n = 22) had two siblings, 90 % (n = 54) had a housewife mother, 56.7 % (n = 34) had a freelancer father, the mothers of 40 % (n = 24) were between 36 and 40 years old and fathers of 31.7 % (n = 19) were between 41 and 45 years old.

It was detected that 78.3 % (n = 47) of the children in the study did not have a cancer history in their families. When disease diagnosis of the children were examined, the most observed three types are that 33.3 % (n = 20) was treated for ALL, 13.3 % (n = 8) for Medullablastoma and 11.7 % (n = 7) for Hodgkin-NonHodgkin Lymphoma. When the counties where the children were living were investigated, it was found that 31.7 % (n = 19) were in Eregli, 31.7 % (n = 19) in Center and 18.3 % (n = 11) in Caycuma. When disease diagnosis was compared based on the county, it was detected that there was a statistical significance (p = 0.016 < 0.05). General distribution map of cancer cases was shown in Fig. 2 and the distribution map of various cancer types were shown in Fig. 3.

Fig. 2
figure 2

Overall distribution map of cancer cases

Full size image
Fig. 3
figure 3figure 3figure 3figure 3

Distribution maps of various cancer types

Full size image

When distribution maps of the children included in the study were examined based on their residence, it was seen that number of cancer cases in Kdz. Eregli and Center were the highest (n = 19).

Discussion

Children constitute a significant group who are affected by physical, biological and social environment. 14 million children smaller that 5 years old die due to unhealthy drinking water, insufficient sanitation, environmental pollution, common diseases and starvation in developing countries each year [12]. Protection of children from environmental chemicals is one of the important problems in modern societies [14]. Incidence of childhood cancers can never be decreased down to zero level, but many of them can be avoided and kept at acceptable levels if sufficient protection measures are taken. Preventable childhood cancers constitute a significant public health issue.

The jobs of mothers and fathers are also important factors in childhood cancers. In the study by Reynolds et al. [18], it was found that the children of fathers who were working with benzene and carbontetrachloride, working in paint and auto repairs and who were exposed to ionizing radiation, electromagnetic field, hydrocarbon and metals and the mothers who were working in beauty centers, textile and dry cleaning jobs were more susceptible to develop leukemia [5]. In a study by Zahm and Ward in China, ALL development risk was found to be 1.8 due to pesticide interaction since their mothers were working in an agriculture-related job, and it was also detected that pesticide interaction during pregnancy has increased ALL risk by 3.5-fold [26]. In our study, mothers of 90 % (n = 54) of children were found to be housewives and fathers of 56.7 % (n = 34) were freelancers; and in contrary to other studies, no statistically significant difference was observed between parents’ occupation and childrens’ disease diagnosis (p > 0.05). However, it can be conclude that both all and ALL cases were accumulated in Center, Kzd. Eregli and Caycuma, the most three industrial areas, in accordance with the findings of Pizzo and Poplack [17].

In the study by Reynolds et al. [18] in California, childhood cancers which were associated with hazardous air pollutants were investigated. In conclusion, it was found that childhood cancers especially leukemias were closely associated with hazardous air pollutants. In the study by Landrigan et al. [15] investigating incidence, prevalence, mortality and cost of chemical substance—related childhood cancers in USA, it was detected that environmental chemical substances were a source of 100 % poisoning, 30 % asthma, 5 % cancer and 10 % neurobehavioral impairment. In addition, the cost of annual poisoning, asthma, childhood cancers and neurobehavioral impairment were calculated as 43.3, 2, 0.3 and 9.2 billion dollars, respectively. This cost was directly medical expenses; it did not include indirect expenses as pain, suffering and late complications. The cost of children disease associated with environmental factors was high, but prevention methods were easier and cheaper [15].

In Environmental Threat to Children Health: Asia Perspective meeting including 15 countries in Philippines in April 2000, it was told that children in Asia were more affected by environmental chemicals. Especially poor children in poor countries were under more risk. Children below five years old are more affected, normal development of these children is not completed due to malnutrition. Safe water is not sufficient; infectious diseases along with contaminated food have a big threat on children. No legal regulations were done against environmental agents in these countries. Education is not adequate and risk awareness was not developed. It is estimated that 25 million people will suffer from cancer and 900 thousand people will have skin cancer in Bangladesh in the future since drinking water is contaminated with arsenic; and most of them are children. Although it was prohibited, hazardous and permanent pesticides, especially DDT are being used in agriculture in Indonesia. In a study performed in Cambodia, 23 % of children between 6 and 12 years old were complaining about skin lesions due to pesticides. Toxic substance interaction of children workers who are working with painting, etc in Asia was increased [3].

Due to the interactions with synthetic chemicals with a high production capacity which were produced in recent 50 years, children are under high risk. Most of these chemicals are commonly used as consumer products and they are distributed around. As a result of environmental pollution with these chemicals, children are affected; because their biological sensitivity are delicate. A comprehensive, national, children-based agenda called Children’s Environmental Health Network-CEHN was formed in order to protect children from environmental diseases. Agenda should be aware about the sensitivity of children against environmental toxic substances and do the following: it should focus on new studies for prevention, should make health risk assessments for children and develop policies for this, start a campaign that train public, healthcare professionals and policymakers about environmental diseases and CEHN should be promoted to establish national environmental health research centers for children for a long-term and fixed investment in order to settle this agenda [14].

In 1979 in USA, the relationship between childhood cancers and high-voltage lines was investigated in 344 children at 18 years old and it was reported that leukemia in children highly increased as proximity of the lines increased. It was also reported that the incidence of childhood cancers increased 2-fold in the houses which were 150 meters near to 200 kilovolt high-voltage line in Sweden in 1982. Positive correlation between electromagnetic fields and cancer cases was revealed in many studies [10].

Similar to the studies mentioned above, higher incidence of cancer cases in Center, Kdz. Eregli and Caycuma counties in our study may be related with the factories in those regions and the chemicals derived from those factories. When some environmental studies held in Zonguldak city were investigated, validity of this proposed thesis becomes stronger. For instance; a research performed by Uyar and Ören [23] in Zonguldak city has revealed that the amount of arsenic was higher than European countries excluding Italy due to Catalagzi Thermal Power Plant (CATES), Eregli Iron and Steel Station (ERDEMIR), the use of coal for heating purposes and traffic-derived emissions [23]. In another study by Uyar et al. [22], amounts of heavy metals in Zonguldak and Kdz. Eregli were detected to be higher than Europe [22]. Polycyclic aromatic hydrocarbons (pah) are significant carcinogens [1]. In a study carried out by Tağıl and Menteşe [20], the relationship between air pollutants (PM10 ve SO2) in Zonguldak and respiratory tract disorders were revealed between 2006 and 2009. Incivez and Bahcelievler districts were detected as cold regions (regions including few cases) and Mesrutiyet and Bastarla districts were detected as hot regions (regions including more number of cases). The reason why more cases were seen in Mesrutiyet and Bastarla districts compared to the others was associated with the geographical locations of these districts. However, they have declared that variability of the number of cases in neighborhoods was associated with variable topography of the region [20]. In a study performed by Çabuk et al. [6], amount of pah in Zonguldak Center and Kdz. Eregli was detected and 6-fold less number of pah were found in Kdz. Eregli compared to Zonguldak. The reason of this is explained by the fact that coal is still used in Zonguldak Center for heating purposes, but natural gas is highly used in Kdz. Eregli instead [6]. Zeydan et al. [27] have determined PM10 concentrations with the help of GIS.

Geospatial distribution and frequency of cancer cases, which are the major health effects of this pollution, should be followed and necessary precautions should be taken by associating them with possible factors. Implementation of such a research is only provided in GIS with geostatistical tests.

In this study, it was first aimed to build up a GIS base which is comprising a numerical county-based map and database, and a detailed geostatistical analysis could not be performed due to the following limitations:

  1. 1.

    Cases were only county-based. A more detailed geospatial distribution map can be obtained if they can be village or district-based.

  2. 2.

    In case of a more detailed detection of the locations of the cases, their relationships with environmental factors (energy transmission lines, industrial institutions, water and sewage system, athmospheric pollutants, etc) will be establisted statistically.

Conclusions

When disease diagnosis was compared based on the county, it was found that there was a statistically significant difference (p = 0.016). However; as it was mentioned in discussion part, GIS-based epidemiological studies do not only provide information about geospatial distribution and temporal spread of the cases, but also they can reveal their relationship with environmental factors geostatistically. However; it is required to know geographical locations of the information regarding the cases as well as the environmental factors in detail. This preliminary study will allow elimination of deficiencies mentioned above and implementation of continuous analyses. Thus, it will be possible to follow CC and take precautions. The future cases must be monitored since one of the main environmental agent caused by usage of coal mines in household heating is being transferred to natural gas providing fresh-air.