Mapping gastrointestinal cancer mortality in Kurdistan province

Introduction

In public health and epidemiologic studies cancer mortality statistics are an extremely important tool for preventing, controlling, effective screening and cure of the disease (1,2). Changes in the disease burdens over the past century indicate that the substantial decrease in the infectious diseases is not specific to developed countries and it is observed in developing countries, as well (3). Cancer is one the five leading causes of death among different age groups of men and women, and following cardiovascular diseases, it is the second cause of death in industrial countries. This is while in Iran, after cardiovascular diseases and car accidents, it is the third leading cause of death. Cancer is the major cause of death among 40- to 60-year-old women and 60- to 79-year-old men of age (4,5). About 1,660,290 new cases of cancer and 580,350 cases of cancer death were expected to occur in the United States in 2013. During the recent 5 years [2005-2009], the delay-adjusted incidence rate of cancer declined by 0.6% per year in men and it was stable in women and the rate of deaths resulting from cancer decreased by 1.8% per year in men and by 1.5% in women. Gastrointestinal cancers include the most common malignant tumors and are among the leading causes of cancer deaths. Reportedly, 290,200 new cases of gastrointestinal cancers and 144,570 such cancer deaths among both genders were expected in the United States in 2013 (6,7).

Drawing geographical patterns of diseases has been used since early twentieth century and it is one of the important tools for determining the spatial pattern of diseases. Utilizing maps is useful both for contagious and non-contagious diseases. Emphasis on the relationship between health and geographical location has a long history in the literature (8,9).

Since gastrointestinal cancers are invasive and lack clear clinical symptoms, most cancer patients who refer to medical centers are at the advanced levels of the disease and have short life spans after the diagnosis (10). Taking this into account and since Kurdistan province is one of the high risk provinces in Iran in terms of gastrointestinal cancer development, we decided to do the current study.

Methods

In this study, digestive system cancer mortality data recorded in rural and urban regions of Kurdistan province between 2005 and 2010 were obtained from the Health and Cure Deputy of Kurdistan University of Medical Sciences and were classified based on the classification in the second chapter of International Classification of Disease, 10^th edition (ICD₁₀). Then the data were fed into SPSS 19 software and were analyzed through appropriate methods. In order to check the relationships between variables, Poisson regression was utilized. Since comparing the mortality rates of populations with different age ranges is complicated, direct and indirect standardization methods were used first. In the indirect method, which is more common, the number of expected deaths can be used but the index which is used more is the ratio of expected deaths to the real deaths which is called standardized mortality ratio (SMR) (11). In order to assess the significance of the difference between the number of deaths in rural and urban areas, the proportion of urban and rural populations were taken into account as the baseline values and Binomial test was used. To calculate the hazard ratios, STATA software was used and to identify high risk and low risk areas, after comparison through indirect method and estimation of SMRs, GIS software was used.

Findings

Out of 3,076 death cases from gastrointestinal cancers, 38% (n=1,168) were women and 62% (n=1,906) were men. Binomial test indicated that the gender ratio in Kurdistan province was significant (0.05). The mean age of the deceased was 68.23±14.69. The ratio of city dwellers to villagers was also significant (0.05). And 55.7% (n=1,713) of deaths occurred in villages and 44.3% (n=1,360) of death occurred in cities and a significant statistical difference was observed between the mean age of the deceased in villages and cities (P<0.05). A total of 246 deaths occurred in Bane, 133 deaths in Sarvabad, 513 deaths in Saghez, 715 deaths in Sanandaj, 467 deaths in Ghorve, 289 deaths in Bijar, 185 deaths in Kamyaran, 236 deaths in Marivan and 29 deaths occurred in other cities.

Mortalities from the gastrointestinal cancers common in Kurdistan province in sequence included stomach cancer with 1,700 cases (55.3%), liver and biliary cancers with 581 cases (18.9%), oesophageal cancer with 402 cases (13.1%), small intestine cancer with 106 cases (6%), colon cancer with 180 cases (5.9%), lips, mouth and throat cancers with 24 cases (0.8%) and rectal and rectosigmoid cancer with 3 cases (0.1%). Most deaths occurred in the seventh decade of life (32.4%). The largest number of gastrointestinal cancer deaths (n=569, 18.5%) were recorded in 2008. The highest rate of gastrointestinal cancer death among women was in Divandare with the rate of 125 per 1,000 deaths and among men it was 121.9 per 1,000 deaths. The rate of gastrointestinal cancer deaths in rural areas equaled 568.91 and in urban areas it equaled 464.64 per 1,000 deaths. Hazard ratio was used to investigate the demographic features in different cities of the province. In all cities except Bane, Divandare and Ghorve, the hazard ratio of men to women was obtained to be more than one. The highest hazard ratio was in Marivan and it equaled 1.392.

As Table 1 displays, the hazard ratio of urban areas to rural areas was obtained more than 1 in Bane and Sarvabad and in other cities, the risk of rural areas was higher. Gastrointestinal cancers in rural areas had a death risk 1.462 times as much as in urban areas and in all cities except Bane and Sarvabad this ratio was significant.

Table 1 Distribution of gastrointestinal cancer deaths by geographical areas
Full table

Since mortality rate is a cumulative value, all independent variables need to be in cumulative form. After fitting the Poisson regression model, the risk of deaths due to digestive system cancers among men was 1.53 times as much as among women and in rural areas it was 1.13 times as much as in urban areas. One of the advantages of Poisson regression is that in these models, the effect of several risk factors can be investigated. The results of such an investigation are shown in the Table 2.

Table 2 Results of the Poisson regression fitting
Full table

As Table 3 shows, stomach cancer with a mortality rate of 110.51 per 100,000, is the most fatal gastrointestinal cancer in the province, and its hazard ratio in men to women is 1.05 and in urban to rural areas is 0.713. Mortality rates from esophageal and colon cancers were significantly different between men and women. Moreover, the mortality rates from stomach, colon and small intestine cancers were also significantly different between rural and urban areas.

Table 3 Mortality rate of cancer by the type them
Full table

In order to determine high and low risk areas of the province, after calculating the SMR through indirect methods, the related map was drawn. Figure 1 displays the observed values of SMRs for gastrointestinal cancer mortality during 2006-2010.

Figure 1 Map of the gastrointestinal cancer deaths across Kurdistan province.

As the map shows, Divandare with an SMR of 1.56 is the highest risk city in the province. A high risk cluster drawn from the southeast to the northwest of the province is observable on the map.

Many studies show that SMRs in less populated regions are a little higher (11). After calculating the density of population in each city and performing the correlation tests, no statistically significant result was observed.

Discussion

In Iran, because of its geographical features and nutritional habits of people, a lot of digestive system cancers are observed in the country and these cancers account for a considerable number of deaths. Prevalence of these cancers is to a great extent different in different parts of the world and among different ethnic groups. A study done in Korea shows that from 1999 to 2008, the annual incidence rate of all cancers increased by 3.1% (12). Stomach cancer is the second leading cause of death from cancerous tumors in the world and the highest percentage of stomach cancer development is reported from China, Korea, and South American countries which may result from different diets (6-11,13-16). This cancer with the mortality rate of 110.51 per 100,000 was found to be the most fatal gastrointestinal cancer in Kurdistan province. Oesophageal cancer forms about 5.5% of malignant tumors of the digestive system and considerable changes have occurred over the past thirty years in its epidemiology in North America and most western countries, and there has been a rapid increase in its incidence (5). Oesophageal cancer with a mortality rate of 26.13 is the third leading cause of death from cancer in Kurdistan. Oesophageal and stomach cancers together as the cancers of upper digestive system are still among the most widespread causes of cancer death in the north of Iran and they are the cause of 16% of annual mortality across the world (14,15). According to a study done by Molanaei and colleagues in 1999 on oesophageal and stomach cancers in Kurdistan province, the risk factors of these cancers were different from those of other regions and alcohol and smoking had a less important role in their incidence (17).

Colorectal cancer with an annual incidence of one million and a mortality of more than 500,000 is regarded a global concern and is the second leading cause of death from cancer. In a study done in 2006 in Kurdistan, periodic outbreak of colorectal cancers during 1995-1999 was about 7 cases per 100,000 populations (18).

As Mohammadpour Tahmtan and colleagues reported, hazard ratio of mortality from digestive system cancers in Mazandarzan and Golestan provinces in men was 55% more than women (8). After fitting the Poisson regression model, the hazard ratio of death from cancer among men was 1.53 times as much as among women and in rural areas it was 1.13 times as much as in urban areas. One reason for this may be more alcohol and tobacco use by men than women. In addition, the hazard ratio in rural areas of all cities except Bane and Sarvabad was higher which may be caused by insufficient awareness to prevent the disease, scarcity of health facilities, hospitals and health care centers in these areas. In the Poisson regression model, the effects of mean age and the ratio of urban to rural areas were compared for the cities of the province. Regardless of urban or rural areas, with each year of increase in the mean age, the gastrointestinal cancer death risk increases by 1.22 times and in equal age conditions, the gastrointestinal cancer death risk in rural areas is 1.12 times more than the urban areas.

Drawing geographical maps is of high importance for decision makers in health arenas of the society for preventing and cure purposes (17). In a study done by Asmarian and colleagues on preparing the geographical maps of the incidence of stomach cancer in Iran over a five-year span, as the result of Kriging regression indicate, Divandare with a variance of 1.01 showed the highest incidence (16.36). Results have shown that the north and northwest of Iran (especially provinces of Ardabil, Mazandaran, and Kurdistan) have a higher incidence of stomach cancer than the south and desert regions (19). The relationships between SMRs and the density of the populations in the cities were not found statistically significant. However, in a similar study, the relationships between SMRs of the road accidents and population densities in Mazandaran province were found to be significant (13). As the geographical map indicates, a high risk cluster draws from the southeast to the northwest areas. Sanandaj and Dehgolan with an SMR of 0.323 were found to be the lowest risk cities in the province. This shows that the levels of health, health facilities and people’s awareness have increased in these cities. High risk cities of Kurdistan province in order were Divandare with an SMR of 1.56, Ghorve with an SMR of 1.09, Saghez with an SMR of 1.07 and Bijar with an SMR of 1.045. The results of the available studies imply the necessity of another study in Divandare to investigate the causes of gastrointestinal cancer development and deaths in this city.

Conclusions

According to the findings of this study Divandare had the most prevalence of gastrointestinal cancers in Kurdistan province which can be related to low education level of habitants in this city so the comprehensive education is recommended.

Acknowledgements

The authors would like to appreciate the personnel of the Health and Cure Deputy of Kurdistan University of Medical Sciences for their cooperation and Research Deputy of Mazandaran University of Medical Sciences for financial support.

Disclosure: The authors declare no conflict of interest.

References

1. Fink AK, German RR, Heron M, et al. Impact of using multiple causes of death codes to compute site-specific, death certificate-based cancer mortality statistics in the United States. Cancer Epidemiol 2012;36:22-8. [PubMed]
2. Weir HK, Thun MJ, Hankey BF, et al. Annual report to the nation on the status of cancer, 1975-2000, featuring the uses of surveillance data for cancer prevention and control. J Natl Cancer Inst 2003;95:1276-99. [PubMed]
3. Sadjadi A, Nouraie M, Mohagheghi MA, et al. Cancer occurrence in Iran in 2002, an international perspective. Asian Pac J Cancer Prev 2005;6:359-63. [PubMed]
4. Koosha A, Mostafa F, Sevil H, et al. Epidemiologic assessment of cancer disease in East Azerbaijan 2007. Med J Tabriz Univ Med Sci Health Serv 2010;32:74-9.
5. Herszényi L, Tulassay Z. Epidemiology of gastrointestinal and liver tumors. Eur Rev Med Pharmacol Sci 2010;14:249-58. [PubMed]
6. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J Clin 2013;63:11-30. [PubMed]
7. Tavoli A, Montazeri A, Mohagheghi MA, et al. Knowledge of cancer diagnosis and quality of life in patients with gastrointestinal cancer. Payesh Journal of The Iranian Institute For Health Sciences Research 2007;6:257-64.
8. Mohammadpour Tahamtan RA, Yazdani J, Zare S, et al. Geographical mapping of gastrointestinal cancer mortality rates in Mazandaran and Golestan provinces, Iran. J Mazand Univ Med Sci 2014;23:125-35.
9. Waller LA, Carlin BP, Xia H, et al. Hierarchical spatio-temporal mapping of disease rates. J Am Stat Assoc 1997;92:607-17.
10. Davood Abadi AH, Sharifi H, Erffan N, et al. An epidemiologic and clinical survey on gastric cancer patients referred to Shahid Beheshti Hospital of Kashan, 1994-2001. J Iran Uni Med Sci 2003;10:211-22.
11. Curtin LR, Klein RJ. Direct standardization (age-adjusted death rates). Healthy People 2000 Stat Notes 1995;(6):1-10.
12. Jung KW, Park S, Kong HJ, et al. Cancer statistics in Korea: incidence, mortality, survival, and prevalence in 2008. Cancer Res Treat 2011;43:1-11. [PubMed]
13. Yazdani Cherati J, Ahmadi Baseri E, Ghadami M. Mapping of Mortality Rate in Suburban Accidents, Mazandaran Province, 2007-2010. J Mazandaran Univ Med Sci 2013;22:50-8.
14. Ghadimi MR, Mahmoodi M, Mohammad K, et al. Comparison of Survival analysis of Gastrointestinal Cancer patients using parametric and Cox models. Journal of School of Public Health And Institute of Public Health Researches 2010;8:1-14.
15. Zarghami N, Mohajeri A, Somi MH, et al. Evaluation of Serum Levels of Zinc and Cupper in Upper Gastrointestinal Cancer Patients and Controls. Med J Tabriz Univ Med Sci Health Serv 2011;33:47-50.
16. Borrelli F, Capasso R, Russo A, et al. Systematic review: green tea and gastrointestinal cancer risk. Aliment Pharmacol Ther 2004;19:497-510. [PubMed]
17. Molanaie N, Yazadanpanah K, Reshadmanesh N, et al. Epidemiology of gastric and oesophageal cancers in Kurdistan province in 1999. Scientific J Kurdistan Uni Med Sci 2000;16:6-9.
18. Molanaie N, Rahimi E, Aiobi S. Epidemiology of colorectal cancer in Kurdistan Province during 1995-99. J Kurdistan Med Sci 2000;17:22-5.
19. Asmarian N, Kavousi A, Salehi M, et al. Mapping of Stomach Cancer Rate in Iran Using Area-to-Area Poisson Kriging. Journal of Health System 2012;8:681-7.