Abstract
Background/Aim: Although the advent of Helicobacter pylori eradication and global societal changes are widely assumed to impact on gastric cancer (GC)-related mortality, there is remarkable little quantitative and qualitative insight into the nature of its effects. Here, we exploited a nationwide reporting system to investigate the epidemiological features of GC-related mortality in China between 2006 and 2013. Patients and Methods: GC mortality data between 2006 and 2013 were obtained from the National Disease Surveillance System published by the China Center for Disease Control and Prevention (CDC). Results: GC mortality increased by 8.2% (from 18.87/100,000 to 20.41/100,000), while GC mortality standardized by the age scale of the population in 2010 decreased by 17.8% (from 21.87/100,000 to 17.98/100,000). Standardized GC mortality in males (25.66/100,000 to 33.89/100,000) was higher compared to females (10.72/100,000 to 14.79/100,000), while standardized GC mortality in rural areas (19.17/100,000 to 26.46/100,000) was higher than in urban areas (15.48/100,000 to 20.04/100,000). Both crude and standardized rates in the 0- to 29-year-old group increased by 22.3% and 16.2%, respectively; while these rates declined in the 30- to 59-year-old group and over 60-year-old group. The proportion of GC deaths that accounts for all cancer deaths declined from 15.99% (2006) to 13.6% (2013); however, the proportion in the 0- to 29-year-old group revealed an increasing trend from 2006 (3.20%) to 2013 (3.87%). Conclusion: Our results reveal a remarkable increase in GC-related mortality in subjects under the age of 30 calling for further measures to prevent the increase in the incidence of GC in young patients.
Gastric cancer (GC) is one of the most common cancers both globally, as well as in China. According to the cancer registry data in the National Cancer Registry Center, the incidence of GC was 36.21/100,000 in 2009 making it the second most prevalent cause of cancer-related death after lung cancer (1). Importantly, pathogenesis of GC is now fairly well-understood and has revealed an importance of Helicobacter pylori infection in its pathogenesis (2). Accordingly, screening, treatment and prevention of H. pylori colonization to reduce the incidence of GC has become the accepted way forward and, in this respect and response, H. pylori colonization levels are universally decreasing (3, 4). The impact, however, of these efforts on GC mortality remains obscure at best.
Other developments may also substantially impact on GC incidence and mortality. In China, a country that substantially contributes to worldwide GC mortality, there has been a huge development of its economy. In turn, this poses severe challenges with respect to control and prevention of GC involving such factors as environmental pollution, the aging of the population and increased psychological pressure. Qualitative and quantitative analyses of the potential influence of such trends on GC mortality are lacking. Earlier epidemiological studies have indicated that the incidence and mortality of GC demonstrates gender and geographic differences (5, 6) but whether this still holds true in changes in society remains largely unexplored.
The considerations mentioned above prompted us to investigate current trends in GC-associated mortality and relate those to gender, geographic and demographic factors. To this end, we exploited the opportunity provided by the data obtained from the National Disease Surveillance Points (DSPs) system of China providing an unusually large population for the analysis of trends in GC-related death. Importantly, we found that although GC-related death, in general, is declining, this does not hold true for subjects of 30 years and younger. Does the data, thus, call for increased efforts to prevent GC in young people?
Patients and Methods
Data sources. In the present study, mortality data was obtained from the DSPs system between 2006 and 2013, which was published by the Chinese Center for Disease Control and Prevention (7-14). The DSP data is primarily based on information from individual counties and then grouped to specific geographic regions with sampling further stratified by urban or rural location and per capita gross domestic product. The DSP system was purposely designed to allow nationwide analysis (15). From 2006 to 2012, 161 surveillance points in the DSPs system were distributed to all 31 provinces (including autonomous regions and municipalities). The total population within the DSPs system was more than 70 million, which was divided according to geographic regions (Eastern, Middle and Western regions; Eastern region: Beijing, Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong and Hainan; Middle region: Shanxi, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hubei and Hunan; Western region: Nei Monggol, Guangxi, Chongqing, Sichuan, Guizhou, Yunnan, Tibet, Shanxi, Gansu, Qinghai, Ningxia and Xinjiang; Figure 1). The sampling strategy and characteristics of this system have been described in detail elsewhere along with quality control measures and procedures for collecting data, coding the cause of death and determining the underlying cause of death (16-18). Following the expansion of the system to 605 surveillance points in 2013, the population monitored increased to over 300 million. This nationwide mortality surveillance system was subsequently established employing same general principles applied for developing the apparently successful DSPs system. It allows confident assessment of both national, as well as provincial, mortality characteristics (19) and its data constitutes a direct extension of the DSP data obtained from 2004 to 2012 and can be considered comparable. Nevertheless, in the present study, before performing data summarization, a quality evaluation was conducted for the data of each surveillance point, in order to eliminate data of surveillance points that exhibited serious omissions in the task report. The final resulting database contained data of 432 monitoring points and was used for final analysis (14). All medical certification information regarding the resident's death for each case was registered in the DSPs system, including deaths in all medical institutions, at home or other places. These registered cases were catalogued by the International Classification of Diseases (ICD-10) in which the code for a malignant tumor and GC is C and C-16, respectively.
Statistical analysis. Descriptive statistical analysis was conducted using SPSS (version 17; SPSS, Inc., Chicago, IL, USA) on relevant data including gender-related death, age-related death, region-related death and mortality. Total population within the National Disease Surveillance Points system was edited by the ArcGIS 10.3 software (http://www.esri.com/arcgis/about-arcgis). The analysis of the time trend and changes in GC mortality was based on data in the DSPs system between 2006 and 2013, while the age of the population was standardized by the population in 2010. Crude and standardized mortality were also described. Age and gender of death distribution were obtained using the DSPs system database in 2013. Fisher's exact test was conducted to analyze gender and urban-rural differences; results were considered statistically significant when p was <0.05.
Results
Trend of GC mortality. It is to be expected that increased screening for H. pylori in conjunction with the rapid societal changes in Chinese society would substantially impact on death associated with GC. Thus, the current study was undertaken to define this influence. Crude GC mortality slightly increased by 8.2% (from 18.87/100,000 in 2006 to 20.41/100,000 in 2013). However, after standardizing the age of the population, standardized GC mortality revealed a descending trend (from 21.87/100,000 to 17.98/100,000, a decrease of 17.8%; Figure 2, Table I). Standardized mortality in males was higher than in females (male: 25.66/100,000 to 33.98/100,000, female: 10.72/100,000 to 26.46/100,000; sex ratio: 2.0 to 2.5). The ratio of GC mortality between the urban and rural areas was between 0.6 and 0.9 (rural area: 19.17/100,000 to 26.46/100,000, urban area: 15.48/100,000 to 20.04/100,000). Figure 3 illustrates the decreasing standardized mortality rates in males, females, urban areas and rural areas with a decrease of 18.1% (from 31.34/100,000 in 2006 to 25.66/100,000 in 2013), 17.6% (from 13.02/100,000 in 2006 to 10.72/100,000 in 2013), 12.8% (from 17.75/100,000 in 2006 to 15.48/100,000 in 2013) and 20.8% (from 24.22/100,000 in 2006 to 19.17/100,000 in 2013), respectively. Crude and standardized GC mortality indicate that mortality rates in the Eastern and Middle areas were slightly higher than in the Western area (Figure 3). Hence, in general, standardized GC mortality decreases in China.
Gender and urban-rural differences in GC death. Based on data obtained from the National Disease Surveillance System, mortality rate was found to be age-specific (Table II) in which mortality due to GC remained zero up to the age of 14. After this age, an increasing trend was found in which mortality started to gradually increase, followed by a more significant increase by the age of 40. This trend was even more evident in male patients. Generally speaking, the GC morality rate of males was higher than that of females. However, before the age of 40, mortality rates of different genders did not reveal any statistically significant difference (p>0.05). In addition, there were no statistically significant differences in mortality rates in the 0-35-year-old and 40- to 44-year-old groups between the rural and urban areas (p>0.05). All the other age groups demonstrated a higher mortality in rural areas than in urban areas. Thus, a disparity between younger and older subjects appears in which, for older subjects, urbanization is an important factor but not for younger subjects.
Geographic regional differences in GC mortality. Based on data obtained from the National Disease Surveillance System, the mortality rate was found to be age-specific (Table III), which remained from zero up to the age of 14. This revealed a higher trend in the 15- to 19-year-old and 25- to 29-year-old age groups in the Western area, compared to the Eastern and Middle areas (p=0.0277 and 0.0099, respectively); mortality rate increased after the age of 40, especially in the Eastern and Middle areas. Mortality in the Eastern and Middle area was significantly higher than in the Western area after the age of 65 (p<0.0001). Thus, again, we observed that younger subjects show a different region-dependent GC-associated mortality when compared to older subjects.
Age group differences in GC mortality. The apparent discordance between younger and older patients with respect to GC-related mortality prompted us to an age-stratified analysis of trends in GC-related death. To this end, patients were divided into three groups by age (0- to 29-year-old, 30- to 59-year-old and 60-year-old and over). Data revealed that both crude and standardized mortality rates increased by 22.3% and 16.2%, respectively, between 2006 and 2013 in the 0- to 29-year-old; while a decrease in crude and standardized mortality rates was revealed in the other two groups. The most dramatic decline in standardized mortality rate was 32.7%, which was found in the 30- to 59-year-old group (Figure 4). Thus, there is a clear trend towards more GC-related death in the recent decade in, especially, young subjects. Indeed, gastric cancer death accounts for up to 13.06% of all malignant tumor-related deaths in 2013, which shows a decrease in malignant tumor-related deaths compared to 2006 (15.99%). The constituent ratio of GC-caused death also revealed a descending trend in both the 30- to 59-year-old and over 60-year-old groups, while this trend increased from 3.20% in 2006 to 3.87% in 2013 in the 0- to 29-year-old group (Figure 4).
Discussion
Many changes in society can be expected to impact on GC-related death, especially increased screening but also the rapid development many societies undergo. At present, there exist little data reported as to how all these developments have affected GC-associated mortality. The present study documents these changes and reveals a relative increase of GC-related death in younger subjects. Thus, our data calls for programs aimed at reducing GC in this population specifically.
This study involved a secondary analysis on the basis of DSPs published over the years. There are unfortunately limitations in the comprehensiveness and timeliness of information. For instance, we can only superficially analyze the conditions on the national death of GC, the incidence rate, the course of the disease, other information on GC, as well as the latest data for 2014 and 2015, which are still lacking. In addition, since the result of the data set is the summary of the death case submitted by each surveillance point, the phenomenon of report omission is inevitable, indicating that our research data are underestimated. However, the DSPs system is currently the most representative and authoritative continuous data system on the cause of death in China. This published data set can reflect the variation trend on the death of various diseases and present the characteristics of the death of different genders, ages and territories. Thus, we feel the trends observed as credible.
According to a Chinese cancer registration report, the incidence of GC in China increased by 12% (from 32.23/100,000 in 2005 to 36.21/100,000 in 2009) (1, 20). In this study, we analyzed GC death data from 2006 to 2013 and our results indicate that crude mortality rate slightly increased in seven years by 8.2%. However, age-standardized mortality revealed a descending trend, which decreased by 17.8% (from 21.78/100,000 to 17.98/100,000). Divergent trends between crude and age-standardized mortality are exactly a representation of an aging population structure under today's rapid developing conditions. Compared to the increasing incidence of GC, the decline in standardized mortality rate is also related to the improvement in medical healthcare. In addition, the development of diagnostic techniques and enhancement of health consciousness have obviously increased the early diagnostic rate of GC. For instance, in Feicheng, Shandong province, China, endoscopy screening test for GC has been listed as an effective method for people older than 40 years (21). Meanwhile, improvement in operation skills, standardization of GC D2 surgery and development of tumor chemotherapy have also played a vital role in decreasing GC mortality (22).
Our study found that the standardized mortality of GC in males (25.66/100,000 to 33.89/100,000) was higher than in females (10.72/100,000 to 14.79/100,000) with a sex ratio of 2.0 to 2.5. In particular, people over the age of 40 have an even higher male predominance in mortality, which is consistent with other previous global reports and studies (23). This phenomenon may be due to the different social roles between males and females in which men usually have higher working pressures, irregular diet, more frequent social activities and other lifestyle problems, such as drinking and smoking (24). An epidemiological set of data in Korea included a total of 3,242 patients with GC between the age of 18 to 45, with all subjects being stratified into three groups based on age (A: 18-30 years; B: 31-40 years; C: 41-45 years). Male-to-female incidence ratio was 1.2:1.0 and male-to-female incidence ratios, based on age groups, was 1.0:1.6 in group A, 1.0:1.0 in group B and 1.7:1.0 in group C, respectively. This indicates the higher incidence of GC in females at younger ages (24). This is consistent with our study in which a higher mortality of GC in females was observed in the 20- to 24-year-old group. This might be related to the critical role of sex hormones, especially the different estrogen levels in young females, in the pathogenesis of gastric cancer (24, 25). Kim et al. found that females with GC were significantly younger and had a different signet ring cell (SRC) histology compared to males; furthermore, females had significantly poorer prognostic factors among young patients (aged ≤45) with SRC (26). Chung et al. further confirmed that excessive exposure to estrogen with a lack of progesterone is a key factor for the development of GS in young females (24).
In our study, we found that both urban and rural standardized mortality decreased by 12.8% and 20.8%, respectively. The standardized mortality of GC in rural areas (19.17/100,000-26.46/100,000) was higher compared to urban areas and the urban-rural proportion fluctuated from 0.6 to 0.9. In age-defined subgroups, we found that there was no statistical difference in mortality in both the 0- to 35- and 40- to 44-year-old groups between the urban and rural areas (p>0.05), while all other age groups demonstrated a higher mortality rate in rural areas than in urban areas. Guo et al. used join point analysis to detect changes in trends and generalized additive models to study birth cohort effect of risk factors between 1987 and 2009. They also found that both urban and rural age-standardized mortality decreased at different time intervals from 1987 to 2009 and in rural areas was higher compared to urban areas in both male and female; however, their study was conducted for all ages and failed to compare data between age groups (27).
This geographically-related difference was also present in correlation to studies conducted in Lithuania, Korea and Brazil (28-30). This rural-urban gap can be mainly explained by different healthcare recourses and prevention methods. H. pylori infection has been identified as the most important risk factor for gastric cancer, which was widely distributed in populations of developing countries, such as Brazil and China (27, 31, 32). In these countries, rural and urban areas are not strictly comparable due to lower economical levels, insufficient healthcare resources, poor living conditions and unhealthy eating habits in rural areas, which are directly related to the higher prevalence of Helicobacter pylori infection (33).
Mortality in Eastern and Middle areas was significantly higher than in Western areas, especially in individuals older than 65 years (p=0.0000). First, this may be due to higher economic levels with seriously aging populations in the Middle and Eastern regions. Secondly, possible dietary factors, such as the salty food, and low consumption of vegetables and fruits have been related to GC development (34, 35). As the majority of the Eastern regions are coastal cities, the corresponding intake of sea products is higher, causing higher intake of salts. A recent meta-analysis indicated that medium or higher intake of salt could increase the incidence of GC, when compared to a lower salt intake with risk ratio (RR) values of 1.68% (95% confidence interval (CI)=1.17-2.41) and 1.41% (95% CI=1.03-1.93), respectively (36). In addition, obesity was also related to an increased risk of early GC (both with respect to cardia and non-cardia cancers). In a recent nationwide case-control study (37), over one-third of Chinese adults were overweight or obese, while in Eastern and Middle major cities like Beijing and Shanghai more than half were overweight or obese (38). As Huang et al. confirmed that an increased body mass index (BMI) was associated with an increased risk of gastric high-grade dysplasia in both men and women and higher serum total cholesterol increased the risk of gastric non-cardiac high-grade dysplasia in women (39), obesity seems important factor explaining GC mortality trends.
Our study revealed an increasing trend of mortality in young patients. From 2006 to 2013, both crude and standardized mortality rates in the 0- to 29-year-old group increased by 22.3% and 16.2%, respectively. Furthermore, the composition of GC-caused death in all malignant tumor-related deaths also increased from 3.20% in 2006 to 3.87% in 2013. Merchant et al. evaluated the trends of gastric cancer incidence rates from 1992 to 2011 by using the Surveillance, Epidemiology and End Results (SEER) registry by calculating the annual percent change (APC) across three age groups (20-49 years, 50-64 years and 65 years or older) and four racial/ethnic groups (Hispanics, non-Hispanic whites, blacks and Asian/Pacific Islanders), they found the APC of the incidence of gastric cancer in young Hispanic men (20-49 years) places it among the top cancers with rising incidence in the USA (40). Several studies have reported more advanced tumors' presentation in young populations. Rona et al. (41) demonstrated that patients with GC in the younger group (YG) (aged ≤45 years) had a higher incidence of stage III/IV disease (86.8% vs. 57.9%, p<0.001), poorly-differentiated carcinoma (95.9% vs. 74.4%, p<0.001) and signet-cell type tumor (88.4% vs. 32.2%, p<0.001) relative to the older group (OG). Furthermore, obesity has been associated with GC (37, 42), with obesity rates rising in the young (43). Wang et al. analyzed data collected from children 7-17 years old from the China Health and Nutrition Survey (CHNS) conducted in 1997, 2000, 2004, 2006, 2009 and 2011. From 2,814 participants with 6,799 observations, the prevalence of overweight and obesity increased in boys and girls between 1997-2011 from 6.5% to 15.5% in boys and from 4.6% to 10.4% in girls (44). It is, thus, tempting to link the specific increase in GC-related death in the young to increases in obesity in this group; obviously, however, further work is necessary to confirm this notion.
Therefore, more studies on GC incidence in young patients needs to be conducted, in order to validate the appropriate age for screening tests in high-risk populations, especially in individuals under the age of 30. Direct measures and strategies aimed at risk factors and treatment improvement should be proposed to inhibit the growth trend of GC in young people.
Acknowledgements
The Authors would like to thank the staff for all their work in collecting and collating the mortality data of gastric cancer obtained from the monitoring death data set in the National Disease Surveillance System published by the China Center for Disease Control and Prevention (CDC).
Footnotes
↵* These Authors contributed equally to this study.
Funding
This work was supported by the National Natural Science Foundation of China (Grant no. 81541050 and 81172317), the Beijing Natural Science Foundation of China (Grant no. 7154191), the Beijing Talent Training Funding of China (Grant no. 2014000021469G266 and 2016000021469G225), the Beijing Municipal Administration of Hospitals Clinical Medicine Development of Special Funding Support (Code: ZYLX201504) and the Beijing Municipal Administration of Hospitals' Youth Programme (Grant no. QML20160105).
Conflicts of Interest
The Authors declare that they have no conflict of interest.
- Received May 19, 2017.
- Revision received June 6, 2017.
- Accepted June 7, 2017.
- Copyright© 2017, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved