Urban-Rural Disparity in Helicobacter Pylori Infection-Related Upper Gastrointestinal Cancer in China and the Decreasing Trend in Parallel with Socioeconomic Development and Urbanization in an Endemic Area.

Background Globally China has the largest urban-rural disparity in socioeconomic development, and the urban-rural difference in upper gastrointestinal cancer (UGIC) is similar to the difference between developed and developing countries. Objectives To describe urban-rural disparity in UGIC and to emphasize prevention by socioeconomic development and urbanization in China. Methods Age-standardized incidence rates (ASRs) of cancers in 2012 were compared between urban Shijiazhuang city and rural Shexian County, and trends from 2000-2015 in Shexian County were analyzed. Findings Compared with urban Shijiazhuang city, the ASR of gastroesophageal cancers in rural Shexian County was 5.3 times higher in men (234.1 vs 44.2/100,000, P < .01) and 9.1 times higher in women (107.7 vs 11.8/100,000, P < .01). This rural-urban disparity in UGIC is associated with differences in socioeconomic development in annual gross domestic product (GDP) per capita of US$2700 vs US$6965, in urbanization rate of 48% vs 100%, and in adult Helicobacter pylori infection prevalence of 75% vs 50%. From 2000-2015, the GDP per capita in Shexian County increased from US$860 to US$3000, urbanization rate increased from 22.4% to 54.8%, and prevalence of H pylori infection among 3- to 10-year-old children decreased from 60% to 46.1% (P < .01). Meanwhile, the biennial ASR of esophagogastric cancer decreased 42% in men, from 313.5 to 182.1 per 100,000 (P < .01), and 57% in women, from 188.6 to 80.4 per 100,000 (P = .00). However, lung, colorectal, and gallbladder cancers and leukemia in both sexes and breast, ovary, thyroid, and kidney cancer in women increased significantly. Despite this offset, ASR of all cancers combined decreased 25% in men (from 378.2 to 283.0/100,000, P = .00) and 19% in women (from 238.5 to 193.6/100,000, P = .00). Conclusions Urban-rural disparity in UGIC is related to inequity in socioeconomic development. Economic growth and urbanization is effective for prevention in endemic regions in China and should be a policy priority.

K E Y W O R D S socioeconomic development, urbanization, urban-rural disparity in cancer, social determinants of health, Helicobacter pylori infection, disadvantaged population.

I N T R O D U C T I O N
In China, the former Soviet Union-style central economy planning system implemented from 1949 until 1980 had invested more in urban industry development than in rural living condition improvement. This has created multidimensional urban-rural inequalities in social determinants of health. According to the National Health Service Survey 2003, safe drinking water is available to 96% of the population of large cities but to <30% in poor rural areas. Ninety percent of residents in large cities have sanitation toilets, compared with <10% in poor rural areas. 1 Although economic reform has lifted millions of people out of extreme poverty in recent decades, the urban-rural income gap has widened, with the rural residents having an annual average per capita disposable income less than a third of urban residents in 2010 (US$898 vs US$2900). 2 These inequities are associated with significant differences in prevalence of Helicobacter pylori infection. As reported by a nationwide collaborative survey in China between 2002-2004, the prevalence of H pylori infection among rural and urban residents was 73.4% vs 65.4% (P < .01). 3 A more recent systematic review reported that the mean prevalence of H pylori infection was 66% for rural Chinese populations and 47% for urban Chinese populations. 4 Because H pylori infection has been established as the most important carcinogen of upper gastrointestinal cancer (UGIC), 5 a huge urban-rural disparity in H pylori infectionrelated cancers has been reported. For example, in China, although the age of rural populations was significantly younger and the population numbered 149.1 million less than the urban population by the middle of 2015 (611,060,000 vs 760,160,000), 6 the numbers of newly diagnosed esophageal and gastric cancer cases in 2015, as estimated by the Chinese National Cancer Registry (CNCR), were 3 and 2 times more among rural than among urban populations (364,100 vs 113,800 and 444,000 vs 235,200, respectively). 7 The urban-rural disparities in these H pylori infectionrelated cancers are significantly larger than that in all cancers combined, which was 213.6 vs 191.5 per 100,000. 7 CNCR also estimated that the 5-year relative survival for new cancer cases diagnosed between 2003 and 2005 in rural China was about half that of their urban counterparts for all cancers com-bined (21.8% vs 39.5%), because of a scarcity of medical resources for early diagnosis and adequate treatment. 8 Despite such grave social inequities, specific policy is lacking. Using age-standardized incidence rates by population-based tumor registration, this paper describes the urban-rural disparity in H pylori infection-related cancers in relation to unfair distribution of social determinants of health and examines the hypothesis that a decreasing trend of these cancers is associated with socioeconomic development and urbanization in an endemic region. Our aim is to emphasize the role of socioeconomic development among disadvantaged populations and make it a policy priority.

M AT E R I A L S A N D M E T H O D S
Shexian County. Shexian County of Hebei Province is located 250 km southwest of Shijiazhuang city in the Taihang Mountain range, from 36 ○ 17' to 36 ○ 55' north latitude (Supplementary Fig. 1 in the online version at doi:10.1016/j.aogh.2017.09.004). Its population was 408,995 in 2012. It is a northern neighbor of Linxian County and a northwestern neighbor of Cixian County. All 3 counties have been noted for endemic rates of esophagogastric cancers. 9 Shexian consists of a county town and 580 villages. The economy of the county town relies on employment in the industry of power generating and steel making, agricultural product refinery, food manufacture, and tourism, and the rural population is engaged in growing rice, wheat, persimmon, Chinese prickly ash, and walnut. A quarter of the Shexian population, mainly young or middle-aged villagers, move between the village and the county town doing seasonal jobs. Diet in Shexian consists of a self-made staple foods of rice or steamed buns flavored by dishes of vegetables, pork, or eggs. Physical activities for villagers mean manual labor. But in the county town, people aged 60 or older walk or jog in the morning or evening.
As in most of China, health care in Shexian has been hospital centered. From 1949 until 2010, only government or public employees were protected by public medical insurance programs. For villagers, only basic medicine was provided free of charge by a rural corporative medical program. Shijiazhuang City. Shijiazhuang is the capital city of Hebei Province, located 270 km south of Peking at the cross point of the north-south-bound Peking-Wuhan railway and the east-west-bound Dezhou-Taiyuan railway, at the foot of Taihang Mountain. The registered urban population in 2012 was 2,374,827. Shijiazhuang was not a city before 1953. It used to consist of several dozen villages. Its urbanization began as late as 1953, when it was chosen by the Chinese central government as the capital city of Hebei Province and as one of the cities with development priority during the first national 5-year program (1953-1957). During implementation of the program, the North China Pharmaceutical Corporation, which is the biggest antibiotic factory in Asia, and 5 modern textile plants were established in Shijiazhuang city. In 60 years, its registered urban population increased from 182,188 in 1953 to 2,374,827 in 2012. In 2012 the gross domestic product (GDP) per capita of Shijiazhuang city was US$6964.80, ranking 119th among the 150 Chinese cities with a GDP higher than the average of US$6102 out of a total of 293 prefecture-level cities. 10

Cancer Registration in Shexian County and
Shijiazhuang City. Population-based cancer registration was established in Shexian County in 1999 and in Shijiazhuang City in 2011, under the agreement of the Chinese National Cancer Registration Center and International Association of Cancer Registration (IACR). The 2 registries are run by the same team according to the principles of the IACR. The Ministry of Health of China provides a running budget. Cancer diagnoses are reported to registries from multiple sources, including local hospitals and community health centers, as well as the Urban Resident Basic Medical Insurance program and the New Rural Cooperative Medical Scheme, both of which are public insurance programs gradually established after the severe acute respiratory syndrome epidemic in 2003 with universal coverage of local registered urban or rural residents achieved by 2010. Data quality of cancer registration is assessed annually by CNCR before publication. On July 21, 2017, the incidence data of Shexian Cancer Registry for 2008-2012 were accepted by IACR for inclusion in the Cancer Incidence in 5 Continents Vol. XI.

Surveillance on the Prevalence of H Pylori Infection among Children Aged 3-10 Years in Shexian County From 2000-2015 as an Indication of Improved Sanitation Associated with Socioeconomic
Development and Urbanization. To examine the effect of improved sanitation associated with socioeconomic development and urbanization between 2000 and 2015 on the prevalence of H pylori infection, we surveyed the prevalence of H pylori infection among a random sample of children aged 3-10 years in 2000 and 2015, which may serve as a more sensitive outcome indicator than adult prevalence. The prevalence of infection was assessed by 13 C urea breath test. Sampling was made by 1/100 sampling from the register book of all Shexian preschool and school children aged 3-10 years edited by the Shexian Department of Education Administration. If the sampled child could not be found or refused, a nearby child was invited. Administration of the 13 C-labeled urea pill, collection of the breath sample, and measurement of the 13 C-labeled CO2 in the collected breath sample were performed by X.D.W., D.G.W., Y.Y., Y.T.C., and G.Y.W. Statistical Methods. Gender-and age-specific cancer incidence rates were calculated, and the Segi standard world population (modified by Doll) was used to calculate age-standardized incidence rates (ASRs). 11 ASRs of major cancers were compared between Shijiazhuang and Shexian and the difference was tested by the approximate method. 12 The trends of biennial ASR of cancer in Shexian between 2000 and 2015 were analyzed using Join Point Analysis 4.2.0.2. 13 The study was approved by the Institutional Ethics Review Board of 4th Hospital of Hebei Medical University.
Although the population age distribution was younger in rural Shexian than in urban Shijiazhuang (Supplementary Fig. 2 in the online version at doi:10.1016/j.aogh.2017.09.004), the former had a crude incidence rate of all cancers combined that was significantly higher in both men (359.5 vs 269.1/ 100,000, P < .01) and women (274.2 vs 207.6/ 100,000, P < .01) ( Table 1). The rural excess in ASR of all cancers combined was even larger, 1.7 times higher in men (325.2 and 187.5, P < .01) and 1.5 times higher in women (230.7 and 150.4, P < .01) ( Table 1). The significantly higher incidence rate of all cancers combined in Shexian County was caused by clustering of esophagogastric cancer, which accounts for 72% and 46% of all cancer in men and women, respectively (Fig. 1). The ASR of esophagogastric cancer alone in Shexian is higher than that of all cancers combined in Shijiazhuang in men (234.1 vs 187.5/100,000), although not in women (107.7 vs 150.4/100,000). Gastric or esophageal cancer has always been the first or the second most common cancer in Shexian. Compared with urban Shijiazhuang, the ASR of stomach cancer in rural Shexian is 5.7 times higher in men (162.0 vs 28.5/100,000, P < .01) and 9.3 times higher in women (71.5 vs 7.7/ 100,000, P < .01); esophageal cancer is 4.6 times higher in men (72.1 vs 15.7/100,000, P < .01) and 8.8 times higher in women (36.2 vs 4.1/100,000, P < .01); liver cancer is 3.2 times higher in women (13.3 vs 4.2/100,000, P < .01). In addition to UGIC, the ASR of cervical cancer is 3 times higher among Shexian than Shijiazhuang women (25.0 vs 8.4/ 100,000, P < .01). All 4 of these are infectionrelated cancers associated with disadvantaged socioeconomic development.  The predominance of UGIC was no longer identified 250 km away in Shijiazhuang city. The proportion of esophagogastric cancer was only 24% in men and 9% in women, contrasting with the corresponding percentage of 72% and 46% in Shexian County. On the contrary, lung and breast cancers became the most commonly diagnosed cancers in men and women in Shijiazhuang city. The ASR of lung cancer in Shijiazhuang city was significantly higher than in rural Shexian in men (46.5 vs 30.0/100,000, P < .01), although not in women (19.0 vs 20.8/100,000), and the ASR of female breast cancer was more than 3 times higher in Shijiazhuang than in Shexian (45.3 vs 14.4/100,000, P < .01). After lung and breast cancer, significant urban excess was also noted with colorectal (22.7 vs 11.8/100,000, P < .01 in men and 15.0 vs 9.3/100,000, P < .05 in women), kidney (6.9 vs 2.2/ 100,000, P < .05 in men and 3.9 vs 0.5/100,000, P < .05 in women), pancreas (3.1 vs 0.1/100,000, P < .01 in men), thyroid (1.9 vs 0.1/100,000, P < .05 in men and 3.9 vs 0.8/100,000, P < .05 in women), prostate (5.3 vs 0.7/ 100,000, P < .01), and ovarian cancers (7.0 vs 2.1/ 100,000, P < .05).

Socioeconomic Development and Urbanization and the Changing Risk of Cancer From 2000-2015 in Rural
Shexian County. Shexian has been a traditional rural agrarian county, but as is the case in most parts of eastern and central China, rapid socioeconomic development and urbanization has been taking place since 1978. From 2000 through 2015, the booming of real estate development has led the economy to rebalance from agricultural and heavy industry toward services and tourism. As a result the annual net income of the Shexian people increased from US$600-US$1500 per person. Overall, the annual GDP per capita increased from US$860-US$3000. 14-17 Along with increasing income and market supply of manufactured food, a diet transition took place, with the staple starch food of steamed buns or rice formerly eaten gradually replaced by increasing proportions of animal food such as saturated fat, meat, eggs, and milk. From 2000-2015, more than 100,000 young or middle-aged villagers had migrated from the villages to live and work in the county town. The proportion of urban residents increased from 22.4% (85,000/380,000) to 54.8% (230,000/420,000). In 16 years the urban population increased by 170%, with an annual increase of 10.7%. Because pipe water, flush toilets, and central sewage treatment systems are available in the county town, public hygiene and environmental sanitation are greatly improved.
In addition to rural-urban migration, in situ urbanization also has been in progress since 1978. Because water has been historically scarce in this mountainous land, a key act of in situ urbanization has been the construction of safe drinking water systems. Since the 1980s, a growing budget has been maintained by the government to improve drinkingwater conditions, by drilling deep wells, often 400-500 meters underground, to tap underground water, or building up water purification systems in underground water resource-scarce areas. Consequently, water conditions for most people living in the mountainous county has been gradually improved: The percentage of people drinking rainwater has been reduced from 80% in the 1970s to 10% in 2010. Meanwhile, the prevalence of H pylori infection by 13 C urea breath test decreased significantly from 60% (240/400) to 46.1% (190/412) (χ 2 = 15.8, P < .001) among children aged 3-10 years.
Meanwhile, the county has experienced rapid population aging. The proportion of people aged 60 or older grew from 7.9% in 2000 to 13.9% in 2015.

D I S C U S S I O N
Socioeconomic status is one of the most important determinants of cancer distribution. 18 In developed countries, because the difference in urban-rural socioeconomic development is limited, the urbanrural disparity in cancer is small. 18 But in China, urban-rural inequity in socioeconomic development is so large that the urban-rural difference in cancer distribution is similar to the difference between developed and developing countries. 7,8,19 In addition, an urban excess in overall cancer incidence is often found in western countries because of a surplus of industrial pollutants, population aging, and a more sedentary life-style 18 ; but in China, the incidence rate of H pylori infection-related cancers associated with disadvantaged socioeconomic conditions are so high among the rural population that their overall cancer burden is higher than in urban population. On this point our finding is in agreement with that reported by Chen et al in 2016. 7 Our study indicates a contrasting cancer pattern between a rural and a nearby urban setting (Fig. 1). In rural Shexian, a predominance of H pylori infection-related gastroesophageal cancers was noted, whereas in urban Shijiazhuang westernizationrelated cancers such as lung, colorectal, and female breast cancer prevail. Because Shexian County is an endemic area for UGIC, the urban-rural difference in esophagogastric cancer in our study is greater than that reported by a Chinese national analysis in 2012 between 74 urban and 119 rural registries. 20 However, the differences in westernization-related cancers are about the same. In the national study the urbanrural difference in colorectal cancer was 23.8 vs 16.1 in men and 17.1 vs 11.6 in women, and our results was 22.7 vs 11.8 in men and 15.0 vs 9.3 in women;   population) in both sexes for stomach, esophagus, and liver cancers and in women for cervix cancer, but a significant positive correlation was found in both sexes for colorectal cancer; in men for lymphoma, bladder, and pancreatic cancers; and in women for breast, uterus, and thyroid cancers and tumors of the central nervous system. 21 Regarding the underlying causes of urban-rural disparity in UGIC, we think inequity in socioeconomic development contributes the most. Although developed from a similar rural agrarian background 60 years ago, with people in both areas being from the Han ethnic group descending from common ancestors living along the Taihang Mountain range for thousands of years, Shexian County is considerably underdeveloped compared with Shijiazhuang city. As shown in Table 3, in 2012 the GDP per capita was 2.6 times higher in urban Shijiazhuang than in Shexian county (US$6964.80 vs US$2700). 10,14-17 Regarding living conditions, all Shijiazhuang people were living in an urban setting, having access to chlorinated pipe water and sanitation toilets, 10 whereas 52% of Shexian people were living in rural villages, drinking unchlorinated well water, without sanitation toilets and sewage systems. [14][15][16][17] The prevalence of H pylori infection was significantly higher in Shexian than in Shijiazhuang city in 2000 (75% vs 42.2%, 22 P < .01). 10,[14][15][16][17] Inequity in social determinants of health between Shexian and Shijiazhuang city has existed since the establishment of the People's Republic of China in 1949 under the Soviet Union-style central economy planning system , which emphasized urban industry investment over on rural living condition improvement, like the 1953-1957 first national 5-year development program choosing Shijiazhuang city as the site to build Asia's largest antibiotic factory and 5 modern textile plants. Under the centralized economic system, urban residents work in public factories or enterprises and received a monthly salary, whereas rural agricultural populations live on grains produced themselves. Preferred development of big cities brought about significant inequalities between large cities and rural counties in the form of employment, income, education, sanitation, health service infrastructure, and so on.
The prevalence of H pylori infection is closely associated with local socioeconomic development and sanitation, and this infection has been classified as a group 1 carcinogen for gastric cancer by the IARC. 5 Recently, viable H pylori has been identified in drinking water supplies by cultural and molecular techniques. 25 Using DNA fingerprint analyses, Karita et al 26 found that most H pylori transmission in Japan is waterborne and that infection is strongly associated with the duration of the history of drinking well water. In Shexian County, because of elevated altitude and other geographic and climate conditions, underground water resources have been scarce from ancient times. People traditionally rely on rainwater collected during the rainy season by "dry wells" or underground pits. The water is stored for use throughout the year. Because sewage systems are unavailable, water in the pits is easily polluted in the rainy season by human and animal waste. The water is infested with microorganisms. In addition, unaccustomed to the smell, people do not add disinfectant to the water. In the Linxian-Cixian-Shexian area, the prevalence of H pylori infection among adults has been reported as high as 80%-90%. 9 Epidemiologically, source of drinking water has been found to be significantly associated with the prevalence of H pylori infection. 9 In 2000 we found the prevalence of H pylori infection among a representative sample of Shexian children aged 3-10 years to be 60% by 13 C urea breathe test, and 50% of children who drank rain-water were infected by the age of 5 years. The agespecific prevalence of H pylori infection was significantly lower in pipe water than in rainwater drinking areas.
The UGIC preventive effects of socioeconomic development and urbanization may require a more integrated approach than reducing H pylori infection alone. Esophageal cancer is much less associated with H pylori infection than is gastric cancer, 5 but the incidence of this cancer also decreased significantly. Zhao et al 28 reported that from 1989-2008, the trends in ASR of esophageal cancer had been falling significantly in rural areas from 65.51 to 47.33 per 100,000 (AAPC = −2.1%, P < .05) in men, and from 36.89 to 25.02 per 100,000 (AAPC = −2.5%, P < .05) in women, but the decrease in urban areas was minimal, only from 14.01 to 13.02 per 100,000 (AAPC = −0.2% P > .05) in men and from 5.72 to 3.91 per 100,000 (AAPC = −1.5%, P > .05) in women.
It is surprising to witness a 42% and a 57% decrease in biennial ASR of esophagogastric cancer in men and women over a 16-year period in Shexian. The war on esophagogastric cancer in the Taihang Mountain high-risk region in China began as early as 1959 when the endemic of hard-swallowing disease attracted the attention of late primer Zhou Enlai. 29 Since then many scientists have visited the Linxian-Cixian-Shexian region to study esophagogastric cancer. Fighting the endemic disease has been placed on the top list of nationally supported championships. When Sino-US relations thawed in 1972 after Chairman Mao and US President Richard Nixon met, scientists from National Cancer Institute of the United States arrived in Linxian to carry out a populationbased randomized nutrition intervention trial. 30 After decades of vigorous study, Chinese scientists concluded that high levels of nitroso compounds in drinking water, fungus infection in food, nutritional deficiencies, and genetic predisposition were implicated. 29 To control the endemic, mass movements such as drying grain and water in the sunshine, marketing of riboflavin-intensified salt, balloon cytology, and endoscopic screening have been successively tried. But never before has a systemic and persistent decrease in the incidence of esophagogastric cancer, similar to what we have just mentioned here, been reported, even in the Sino-US joint populationbased intervention trial. 30 This example indicates that assigning causation of poverty-related UGIC solely to biological anomalies has not been wholly successful in accounting for all of the relevant factors in the pathogenesis and progression, because it has been the disadvantaged social conditions that predisposed the population to the onset of pathogenesis. In this sense, social inequalities are fundamental causes of the cancer. Social determinants of health can never be ignored or regarded as secondary to biological factors.
The effect of socioeconomic development is most pronounced when the population is severely disadvantaged. It is perhaps largely because of the disadvantaged socioeconomic status along the Taihang Mountain range in China that we got this opportunity to witness the strength of this connection. Indeed, rural populations in China have been severely and are continually being disadvantaged. The Chinese National Tumor Prevention and Control Office conducted a nationwide cancer mortality survey in 1974-1976. 29 It identified several dozen of rural areas with extraordinarily high mortality rates of specific cancer, such as the Linxian-Cixian-Shexian region for esophageal cancer, Qidong County in Jiangsu province for liver cancer, and Qingan County in Jiangxi province for cervical cancer, and so on. This is the origin of so-called high-risk cancer regions in China. Most of these high-risk regions are located in remote or geographically disadvantaged mountainous areas with low levels of socioeconomic development, and the prevalent cancer is often infection-based, poverty-related cancer. During the 1970s, when chemical carcinogenesis theory was popular, strenuous efforts were made by scholars to identify chemical carcinogens in the local environment, and mass movements were launched to curb the suspected carcinogens. However, cancer rates of these high-risk regions had little or only moderate reduction. Only after rapid socioeconomic development began in the 1980s did significant reductions begin to appear, suggesting that cancer control among marginal groups depends on social progress as a whole. This is particularly relevant to a country as vast and unevenly developed as China.
The effect of socioeconomic development on UGIC is sustainable. Even after a decade of progressive decline, the rural excess of esophagogastric cancer in 2012 in Shexian vs Shijiazhuang was still as large as 234.1 vs 44.2 per 100,000 in men (P < .01) and 107.7 vs 11.8 per 100,000 in women (P < .01), and this disparity was accompanied by a difference in GDP per capita of US$2700 vs US$6965, in an urbanization rate of 48% vs 100%, and in an adult H pylori infection prevalence of 75% vs 50% (Table 3).
These local figures are in agreement with national data. According to the World Factbook published by the US Central Intelligence Agency, even as of 2015, 7% and 36.3% of rural population in China had unimproved drinking water and sanitation. 30 Although 55.6% of Chinese people are living in urban cities, people living in rural villages located far from cities generally have no access to chlorinated pipe water and sanitation toilets. 31 Therefore, there is a long road ahead of the Chinese government to reduce the urban-rural gap in distribution of socioeconomic determinants of health. If this goal is achieved, the socalled high-risk cancer regions may gradually disappear.
Reducing urban-rural inequality in social determinants of health can achieve more health outcomes than cancer in China. A 2006 review by the China Ministry of Health and several United Nations agencies found large inequalities in maternal and child mortality rates between the poorest rural counties and urban areas in 2004, 31 and more recent data indicate that the maternal mortality ratio in 2006 and newborn mortality rate in 2007-2008 in the poorest rural type IV counties both remained around 5 times higher than in urban areas. 32,33 Therefore, government commitments to social equity are bound to achieve extensive health outcomes.
Our study has a few merits. First, previous studies on urban-rural disparity in cancer in China often defined urban or rural population according to the administration level. 32 If it is from a prefecturelevel city (eg, province-governing city) or a municipality, the population is considered urban; if it is from a county or county-level city, the registry is regarded as rural. However, in addition to the urban population, a typical Chinese prefecture-level city also includes several rural counties. When these people are misclassified as urban, the real urban-rural difference is diluted. This may explain the contradictory results previously reported in China. 7,32 Regarding this, we compared the ASR between a pure urban and a rural county. Second, although previous studies have noted declining trends in UGIC in parallel with socioeconomic development in Chinese highrisk areas, 27,28 synchronous increasing trends of urbanization-related cancer are rarely reported. Conversely, previous studies on cancer transition in developed countries usually focus on the increase in westernization-related cancer, 33,34 but the decrease in UGIC was less described, because such observations occur only among disadvantaged population in endemic regions. Along with socioeconomic development and urbanization in an endemic area, we noted contrasting trends of dramatically decreasing UGIC as opposed to steadily increasing westernization-related cancers over a 16-year period. Our study reveals a more realistic view regarding the effect of socioeconomic development and urbanization on cancer transition. Third, in Shexian County, cancer screening programs have never been implemented for lung, colorectal, female breast, and cervical cancers. With UGIC, <5% of the incident cases may be discovered by endoscopic screening. Therefore, an increasing trend in the incidence of westernizationrelated cancers and a decreasing trend in UGICs should not have been influenced by screening.
Our study has a limitation in study design. Ideally, urban-rural disparity in cancer incidence in Shexian should be assessed by comparing the ASR of the county town to that of the 580 villages, rather than choosing Shijiazhuang as the urban site for comparison, because the population would be more similar but with different urbanization rates. However, as we described earlier, socioeconomic disparity in China exists mainly between people in prefecture cities (province-governing cities) or higher-level cities and people in rural counties. Regarding urban-rural disparity between the county town and villages within a county, because the county town is usually not assigned a development priority under the central planning economic system and there are few factories or enterprises located there, the number of nonagricultural salary-receiving people is very small. Only government employees or workers in public sectors are paid with a monthly salary. For example, the total number of salary-receiving people and their families accounts for only 22% of the population (44,000/196,318) living in the county town of Shexian. The remaining 78% used to be registered rural residents. They have migrated into the county town in recent years. These "floating" people had been of agrarian origin and were still counted as rural residents in the Chinese National Census. We compared the ASR of cancers between nonagricultural and agricultural Shexian people for the time between 2008-2012, As shown in Figure 4, although the ASRs of westernization-related cancers are higher among nonagricultural than among agricultural people, with significant differences noted in women for breast, colorectal, lung, kidney, thyroid, pancreas, ovarian, and corpus uterine cancers, and in men for lung, colorectal, kidney, bladder, larynx, and thyroid cancers and brain tumors, the ASRs of UGIC cancers among agricultural people were not significantly higher than among nonagricultural populations, except for male liver and female esophageal cancer, suggesting agricultural and nonagricultural populations in Shexian County have had similar levels of risk for UGIC under a common macroenvironment.
Another limitation may related to the representativeness of our findings. Because Shexian County has been an endemic region for esophagogastric cancers, the rural excess of UGIC may be much greater than that in other places in China. However, as described in the Introduction, although the age of the national rural population was significantly younger and the population size was considerably smaller than their national urban counterpart, 6 the newly diagnosed esophageal and gastric cancer cases were 3 and 2 times more common, respectively, among rural than among urban population in 2015. 7 In this aspect our estimate is representative of the rural excess of UGIC in China.
In summary, by noting urban-rural disparity in UGIC in relation to unfair distribution of social determinants of health, and the trends of deceasing UGIC in parallel with rapid socioeconomic development and urbanization, this paper hopes to call the attention of policymakers to the extraordinary urbanrural disparity in cancer in China and stress the role of socioeconomic development in tackling the rural excess of H pylori infection-related cancers.

S U P P L E M E N T A R Y D AT A
Supplementary figures accompaying this article can be found in the online version at doi:10.1016/ j.aogh.2017.09.004.