Abstract
Objective
Smoking and alcohol consumption explain little of the risk for upper-gastrointestinal (UGI) cancer in China, where over half of all cases in the world occur.
Methods
We evaluated questionnaire-based risk factors for UGI cancers in a case-control study from Shanxi Province, China, including 600 esophageal squamous cell carcinomas (ESCC), 599 gastric cardia adenocarcinomas (GCA), 316 gastric noncardia adenocarcinomas (GNCA), and 1514 age- and gender-matched controls.
Results
Ever smoking and ever use of any alcohol were not associated with risk of UGI cancer; only modest associations were observed between ESCC risk and highest cumulative smoking exposure, as well as GNCA risk and beer drinking. While several associations were noted for socioeconomic and some dietary variables with one or two UGI cancers, the strongest and most consistent relations for all three individual UGI cancers were observed for consumption of scalding hot foods (risk increased 150% to 219% for daily vs never users) and fresh vegetables and fruits (risk decreased 48% to 70% for vegetables and 46% to 68% for fruits, respectively, for high vs low quartiles).
Conclusion
This study confirms the minor role of tobacco and alcohol in UGI cancers in this region, and highlights thermal damage as a leading etiologic factor.
Keywords: smoking, alcohol, socioeconomic status, diet
Introduction
Upper gastrointestinal (UGI) cancers, including esophageal and gastric cancer, are among the most common causes of cancer death in the world, with an estimated 562 000 esophageal and 883 000 gastric cancer deaths annually[1]. Shanxi Province in north central China has among the highest esophageal cancer rates in the world [2]. However, to date, no single dominant environmental risk factor has been identified for esophageal or gastric cancer in this region.
Tobacco smoking is consistently reported as an important risk factor of esophageal cancer [3; 4; 5; 6; 7], especially for squamous cell carcinoma, and gastric cancer [4; 8; 9] in the West, but appears to incur little or no effect for these cancers in Asian countries [9; 10; 11; 12]. Our group previously reported that tobacco smoking was only a modest risk factor for esophageal cancer, and did not increase risk for gastric cardia adenocarcinoma (GCA) or gastric noncardia adenocarcinoma (GNCA) in Linxian, Henan Province[9], which is geographically adjacent to Shanxi Province.
Similar to smoking, alcohol consumption is considered a dominant risk factor for esophageal squamous cell carcinoma (ESCC) in the West [13; 14; 15], but has not been associated with risk in Linxian or Shanxi Province [9; 16].
Among different lifestyle factors, dietary factors may play an important role in the carcinogenesis of esophageal cancer. Some specific dietary elements, like pickled vegetable juice and moldy food, have been reported to be associated with esophageal cancer risk in north China [9; 16]. Nutrition-related deficiencies in vitamins, minerals, and other micronutrients have also been linked to the elevated risk of esophageal and gastric cancer in these areas [17; 18].
To explore the role of these risk factors in esophageal cancer in detail, we conducted a large case-control study in Shanxi Province. Because cancers of the esophagus and stomach are anatomically adjacent, may share common etiologies, and occur at very high rates in this region of north central China, we also evaluated risk factors for gastric cancer, including GCA and GNCA.
Materials and Methods
Study population
This study was based on a case-control study of upper gastrointestinal (UGI) tract cancers conducted in Shanxi Province, China[19; 20]. Briefly, the recruiting criteria for cases include: 1. Males or females over 20 years old; 2. Residents from Taiyuan, Linfen, Jinzhong, Changzhi, and Xinzhou; 3. Recently diagnosed for cancer of the esophagus or stomach without previous treatment; 4. Had surgical treatment for tumor at the Shanxi Cancer Hospital; 5. Diagnoses were histologically confirmed by pathologists at the Shanxi Cancer Hospital and the National Cancer Institute in the United States. Esophageal cancer cases were limited to esophageal squamous cell carcinoma (ESCC), which is the dominant type in Shanxi Province. Gastric adenocarcinoma (GCA) included gastric cancers located in the top three centimeters of the stomach, while gastric noncardia adenocarcinomas (GNCA) were gastric cancers located in the remainder of the stomach. One control was recruited for each case matched on age (± 5 years), gender, and neighborhood of residence. Interviews for controls were completed within six months of matched cases. To identify potential controls, each case was asked to identify a neighbor of approximately the same age and gender. When the initial suggested neighbor could not be enrolled (ie, unavailable, ineligible, or refused), that neighbor, other neighbors, or the village doctor were asked to suggest another neighbor of the same age and gender. Potential controls were asked if they had any cancer or UGI disease, and were considered ineligible if they reported affirmatively to either question. Ninety-five percent of available and eligible controls (ie, the ones actually invited) were enrolled. The primary reason for non-enrollment among available/eligible controls was refusal to give a blood sample.
All subjects signed informed consents and all study interviews were conducted by trained study nurses using a standardized, interviewer-administered questionnaire. Cases were interviewed in the hospital at the time of their admission for the diagnosis and treatment of their cancer (but before their surgery), while controls were interviewed in their homes within six months of the diagnosis of their matched control. Interviews obtained information on demographics and lifestyle during the most of the lives before they became ill for cases and before their interview for controls. Questionnaire-based information on tobacco smoking (cigarettes per day, tobacco amount [grams] per month for pipe smoking, age started and quit smoking); frequency (ie, daily, weekly, monthly, seldom, not at all) of alcohol use (ie, beer, wine, and liquor use separately); socioeconomic status (SES) (occupation for most of life, education, household size, household income, and refrigerator usage); and frequency (ie, daily, weekly, monthly, seldom, not at all) of dietary intake before illness (scalding hot food, moldy food, vinegar, pickled vegetable or pickled vegetable juice, pork, beef, lamb, fish, chicken, eggs, 16 kinds of vegetables, and 12 kinds of fruits) were collected. To capture the impact of the Chinese economic reformation in the late 1970s on food and drink consumption, we asked about frequency of alcohol and dietary intake before and after 1984. For beer, consumption frequency in summer and winter was also asked.
This study was approved by the institutional review boards of the Shanxi Cancer Hospital in Taiyuan, Shanxi Province, China, and the National Cancer Institute in Bethesda, Maryland, USA.
Statistical analysis
Odds ratios (OR) and 95% confidence intervals (CI) were calculated from unconditional logistic regression models adjusted for the three matching factors, including age (continuous), gender, and geographic region (as a proxy of neighborhood). To increase power, all 1514 controls were analyzed together.
As very few women ever smoked (10%) either cigarettes or a pipe, analysis of smoking was restricted to males. Since only 3% of controls were exclusively pipe smokers, we combined these tobacco exposures in most of our analysis; intensity of pipe smoking (grams/month) was expressed as cigarettes/day equivalents (1g tobacco = 1 cigarette), and cumulative lifetime tobacco exposure (pack-year equivalents) was calculated as the product of intensity (packs/day) and duration (years). Intake frequencies of meat and legumes were combined when the number of subjects in frequency categories was small. Consumptions of red meat, chicken, fish, eggs, and soy products were adjusted for each other to control for the effect of substitution; similarly, consumptions of corn, millet, wheat, rice, and sorghum were adjusted for each other. The frequencies of total vegetable and total fruit consumption were calculated as the sum of the number of times per year each food item was eaten. As very few women reported ever drinking any alcohol (8.6%), analysis of alcohol consumption was restricted to males.
For exposure variables that showed significant relations with cancer in analyses controlled for matching factors only, we further assessed their independence in analyses adjusted for additional potential confounders (ie, occupation, education, household size, household income, refrigerator usage, ever smoking, and ever alcohol drinking). Since results of these additional analyses did not alter the magnitude or significance of the basic findings, we only report results adjusted for matching factors here. All P values were two-sided and a P value < 0.05 was considered statistically significant. We used SAS 9.1 for all statistical analyses.
Results
A total of 600 ESCC, 599 GCA, and 316 GNCA cases and their 1514 matched controls were included in the current analyses. Among ESCC cancers, 7% were anatomically located in the upper third of the esophagus, 70% in the middle, and 23% in the lower third. Gender, age, and geographic region distribution of study subjects are shown in Table 1. Overall, nearly three-quarters of cases were males and the median age of all cases was 59 years.
Table 1.
Demographic characteristics of subjects
| ESCC | GCA | GNCA | |||
|---|---|---|---|---|---|
| Controls | Cases | Cases | Cases | ||
| N =1514 | N=600 | N=599 | N=316 | ||
| Age * | All | 59(52-65) | 58(51-64) | 61(55-66) | 57.5(50-63) |
| Male | 60(53-65) | 59 (52-64) | 61.5(55-66) | 58(51-63) | |
| Female | 57(50-63) | 57(50.5-63) | 60(54-64) | 54(44-63) | |
| Gender | Male (%) | 1107(73) | 376(63) | 492(82) | 239(76) |
| Female (%) | 407(27) | 224(37) | 107(18) | 77(24) | |
| Geographic regions | |||||
| Taiyuan (%) | 524(35) | 212(35) | 199(33) | 113(36) | |
| Linfen (%) | 266(18) | 94(16) | 118(20) | 54(17) | |
| Jinzhong (%) | 294(19) | 153(26) | 105(18) | 37(12) | |
| Changzhi (%) | 274(18) | 90(15) | 121(20) | 63(20) | |
| Xinzhou (%) | 156(10) | 51(8) | 56(9) | 49(15) | |
Median (inter quartile range)
ESCC: esophageal squamous cell carcinoma
GCA: gastric cardia adenocarcinoma
GNCA: gastric noncardia adenocarcinoma
Tobacco (cigarettes and pipe) smoking among males
Among the 85% of male controls who ever smoked tobacco, 19% smoked both cigarettes and pipes, 3% smoked pipes only, and 63% smoked cigarettes only (Table 2). Ever tobacco smoking (vs. never) was not associated with risk of any of the three UGI cancers studied here. Based on small numbers (9% of controls and 5% of cases), current pipe smoking was associated with decreased risk of all three UGI cancers. Neither former nor current cigarette smoking status was associated with UGI cancer risk. Compared with never smokers, higher quartiles of cumulative lifetime exposure (pack-year equivalents) were associated with a monotonically increased risk for ESCC (P for trend=0.009), and men with over 40 pack-year smoking exposure had a 1.5-fold increased risk (OR=1.51, 95% CI = 1.01-2.24
Table 2.
Tobacco (cigarette + pipe) smoking and risk for upper gastrointestinal cancer (males only) (cigarette equivalents)
| ESCC | GCA | GNCA | ||||||
|---|---|---|---|---|---|---|---|---|
| Control (%) | Case (%) | OR (95% CI)# | Case (%) | OR (95% CI)# | Case (%) | OR (95% CI)# | ||
| Ever tobacco | ||||||||
| Never | 169(15) | 48(13) | 1 | 93(19) | 1 | 32(13) | 1 | |
| Ever | 938(85) | 328(87) | 1.19(0.84-1.69) | 399(81) | 0.80(0.60-1.06) | 207(87) | 1.12(0.74-1.69) | |
| Ever tobacco | ||||||||
| Never | 169(15) | 48(13) | 1 | 93(19) | 1 | 32(13) | 1 | |
| Cigarettes only | 696(63) | 249(66) | 1.22(0.85-1.73) | 303(62) | 0.83(0.62-1.10) | 162(68) | 1.18(0.78-1.80) | |
| Pipes only | 37(3) | 5(1) | 0.53(0.19-1.47) | 13(3) | 0.60(0.30-1.23) | 5(2) | 0.57(0.20-1.62) | |
| Cigarettes and pipes |
205(19) | 74(20) | 1.22(0.80-1.86) | 83(17) | 0.74(0.52-1.08) | 40(17) | 1.004(0.60-1.68) | |
| Tobacco smoking status | ||||||||
| Never | 169(16) | 48(13) | 1 | 93(22) | 1 | 32(15) | 1 | |
| Former a | 165(15) | 64(17) | 1.37(0.89-2.12) | 92(19) | 0.99(0.69-1.42) | 41(18) | 1.47(0.88-2.48) | |
| Current | 742(69) | 256(70) | 1.17(0.82-1.67) | 289(61) | 0.74(0.56-0.998) | 161(69) | 1.05(0.69-1.60) | |
| Cigarette smoking status | ||||||||
| Never | 206(19) | 53(14) | 1 | 106(21) | 1 | 37(15) | 1 | |
| Former a | 193(17) | 72(19) | 1.43(0.94-2.15) | 107(22) | 1.06(0.76-1.49) | 45(19) | 1.51(0.92-2.46) | |
| Current | 708(64) | 251(67) | 1.28(0.91-1.80) | 279(57) | 0.81(0.61-1.06) | 157(66) | 1.18(0.80-1.76) | |
| Pipe smoking status | ||||||||
| Never | 865(78) | 297(79) | 1 | 396(80) | 1 | 194(81) | 1 | |
| Former a | 144(13) | 62(16) | 1.21(0.87-1.69) | 73(15) | 1.10(0.80-1.50) | 33(14) | 1.08(0.71-1.64) | |
| Current | 98(9) | 17(5) | 0.54(0.30-0.94) | 23(5) | 0.45(0.27-0.74) | 12(5) | 0.45(0.23-0.87) | |
|
Tobacco intensity b (cigarettes/day) |
||||||||
| Never smoker | 169(17) | 48(16) | 1 | 93(22) | 1 | 32(15) | 1 | |
| 1-9 | 253(25) | 39(13) | 0.56(0.35-0.89) | 53(13) | 0.39(0.26-0.57) | 42(20) | 0.84(0.50-1.30) | |
| 10-19 | 184(18) | 55(18) | 1.04(0.67-1.62) | 77(19) | 0.80(0.55-1.15) | 39(19) | 1.08(0.64-1.80) | |
| 20-29 | 324(32) | 135(44) | 1.40(0.96-2.06) | 152(37) | 0.90(0.65-1.25) | 75(36) | 1.17(0.74-1.85) | |
| 30-70 | 77(8) | 30(10) | 1.28(0.75-2.19) | 37(9) | 0.96(0.60-1.53) | 20(10) | 1.33(0.71-2.48) | |
| P for trend | <0.01 | 0.13 | 0.14 | |||||
|
Tobacco duration b (years) |
||||||||
| Never smoker | 169(17) | 48(16) | 1 | 93(22) | 1 | 32(15) | 1 | |
| 1-26 | 203(20) | 66(22) | 1.08(0.68-1.70) | 64(16) | 0.70(0.47-1.05) | 54(26) | 1.22(0.73-2.05) | |
| 27-33 | 192(19) | 69(22) | 1.20(0.78-1.85) | 89(22) | 0.97(0.67-1.41) | 46(22) | 1.20(0.73-2.00) | |
| 34-41 | 233(23) | 72(23) | 1.10(0.72-1.66) | 101(24) | 0.81(0.57-1.15) | 52(25) | 1.18(0.73-1.92) | |
| 42-58 | 206(20) | 52(17) | 0.92(0.58-1.45) | 65(16) | 0.51(0.35-0.76) | 23(11) | 0.65(0.36-1.17) | |
| P for trend | 0.89 | 0.01 | 0.38 | |||||
|
Total tobacco exposure c (pack-years) |
||||||||
| Never smoker | 169(17) | 48(16) | 1 | 93(22) | 1 | 32(15) | 1 | |
| 0.5-14 | 209(21) | 43(14) | 0.75(0.48-1.18) | 73(18) | 0.73(0.51-1.05) | 52(25) | 1.24(0.77-2.02) | |
| 15-26 | 217(22) | 65(21) | 1.09(0.72-1.65) | 72(17) | 0.74(0.52-1.06) | 39(19) | 0.88(0.53-1.45) | |
| 27-39 | 216(22) | 82(27) | 1.39(0.93-2.08) | 90(22) | 0.90(0.64-1.27) | 44(21) | 1.14(0.70-1.86) | |
| 40-157 | 192(19) | 69(22) | 1.51(1.01-2.24) | 84(20) | 0.84(0.60-1.18) | 40(19) | 1.25(0.76-2.04) | |
| P for trend | <0.01 | 0.7769 | 0.5515 | |||||
Adjusted for age (continuous), geographic region (5 classes); OR: odds ratio; CI: confidence interval
Formers smokers were defined as persons who stopped smoking for at least 2 years
Intensity (average usual tobacco per day, cigarette equivalent), and duration (years of tobacco) among smokers
Total exposure was calculated as the product of intensity and duration (pack-year equivalent); 1 pack-year equals 20 cigarettes per day for 1 year
ESCC: esophageal squamous cell carcinoma
GCA: gastric cardia adenocarcinoma
GNCA: gastric noncardia adenocarcinoma
Alcohol consumption among males
Among male controls, about 66% reported ever use of any alcohol, including 43% ever consumption of beer, 23% ever consumption of wine, and 63% ever consumption of liquor (Table 3). Beer was consumed more frequently in summer that winter, while both beer and liquor were consumed by more people after 1984 than before. Ever consumption of any alcohol (versus never) was not associated with risk at any UGI cancer site. Persons who reported ever consuming beer (vs. never) had higher risk for GNCA (OR=1.53, 95% CI = 1.14-2.05); this association was consistent for consumption in both summer and winter, and both before 1984 and after. Wine usage did not show any association with the risk of UGI cancer. Ever use of liquor was unassociated with risk of UGI cancer, but analysis of risk by frequency showed that weekly usage of liquor was associated with increased risk of GCA, for use both before 1984 (OR=2.39, 95% CI=1.30-4.40) and after 1984 (OR=1.64, 95% CI=1.02-2.65). However, this one usage group with elevated risks was based on small numbers (2% of controls versus 4% of GCAs before 1984; 5% of controls versus 7% of GCAs after 1984), and no dose-response for increasing frequency was observed overall.
Table 3.
Alcohol consumption and risk for upper gastrointestinal cancer (males only)
| ESCC | GCA | GNCA | ||||||
|---|---|---|---|---|---|---|---|---|
| Control (%) | Case (%) | OR (95% CI)# | Case (%) | OR (95% CI)# | Case (%) | OR (95% CI)# | ||
| Ever alcohol | Ever vs. never | 731(66) | 270(72) | 1.23(0.95-1.60) | 319(65) | 1.01(0.80-1.27) | 172(72) | 1.29(0.94-1.76) |
| Beer | ||||||||
| <1984 summer | Any vs. never | 299(27) | 119(32) | 1.17(0.90-1.52) | 135(27) | 1.10(0.86-1.40) | 90(38) | 1.58(1.17-2.14) |
| <1984 winter | Any vs. never | 169(15) | 67(18) | 1.15(0.84-1.57) | 89(18) | 1.30(0.98-1.73) | 57(24) | 1.65(1.16-2.33) |
| >1984 summer | Any vs. never | 465(42) | 185(49) | 1.25(0.98-1.60) | 203(41) | 1.06(0.85-1.33) | 129(54) | 1.56(1.16-2.09) |
| >1984 winter | Any vs. never | 181(16) | 72(19) | 1.16(0.85-1.57) | 92(19) | 1.25(0.95-1.66) | 64(27) | 1.74(1.25-2.44) |
| Ever beer | Ever vs. never | 474(43) | 186(49) | 1.23(0.96-1.56) | 206(42) | 1.05(0.84-1.32) | 131(55) | 1.53(1.14-2.05) |
| Wine | ||||||||
| <1984 | Any vs. never | 221(20) | 83(22) | 1.05(0.78-1.40) | 101(20) | 1.10(0.84-1.45) | 59(25) | 1.31(0.93-1.84) |
| >1984 | Any vs. never | 251(23) | 98(26) | 1.11(0.84-1.46) | 112(23) | 1.07(0.83-1.39) | 63(26) | 1.23(0.88-1.71) |
| Ever wine | Ever vs. never | 253(23) | 99(26) | 1.11(0.84-1.46) | 112(23) | 1.06(0.82-1.37) | 63(26) | 1.22(0.87-1.69) |
| Liquor | ||||||||
| <1984 | Never | 533(48) | 160(42) | 1 | 234(47) | 1 | 102(43) | 1 |
| <Weekly | 456(41) | 172(46) | 1.19(0.92-1.53) | 197(40) | 1.03(0.82-1.30) | 108(45) | 1.24(0.92-1.68) | |
| weekly | 23(2) | 14(4) | 1.90(0.95-3.81) | 22(4) | 2.39(1.30-4.40) | 9(4) | 1.95(0.87-4.36) | |
| Daily | 95(8) | 30(8) | 0.99(0.63-1.56) | 39(8) | 0.95(0.63-1.43) | 20(8) | 1.23(0.72-2.12) | |
| >1984 | Never | 421(38) | 121(32) | 1 | 194(39) | 1 | 74(31) | 1 |
| <Weekly | 421(38) | 156(42) | 1.23(0.94-1.63) | 177(36) | 0.96(0.75-1.24) | 102(43) | 1.34(0.96-1.86) | |
| weekly | 51(5) | 20(5) | 1.26(0.71-2.21) | 33(7) | 1.64(1.02-2.65) | 19(8) | 1.88(1.04-3.40) | |
| Daily | 214(19) | 79(21) | 1.18(0.84-1.65) | 88(18) | 0.94(0.69-1.28) | 44(18) | 1.23(0.81-1.88) | |
| Ever liquor | Ever vs. never | 700(63) | 258(69) | 1.20(0.93-1.55) | 306(62) | 1.01(0.81-1.27) | 166(69) | 1.30(0.96-1.77) |
Adjusted for age (continuous), geographic region (5 classes); OR: odds ratio; CI: confidence interval
ESCC: esophageal squamous cell carcinoma
GCA: gastric cardia adenocarcinoma
GNCA: gastric noncardia adenocarcinoma
Socioeconomic status
Compared to farmers, office workers had 50% and 130% increased risks for GCA and GNCA, respectively (Table 4). Persons with the highest level of education (more than 10 years) showed increased risk at all three cancer sites, reaching significant levels for GCA (OR=1.98, 95% CI=1.49-2.64) and GNCA (OR= 1.50, 95% CI=1.05-2.13). Larger household size (more than 4 family members) also appeared to increase risk for ESCC (OR=1.29, 95% CI=1.06-1.58) and GCA (OR=1.31, 95% CI=1.07-1.59). Longer refrigerator ownership was associated with decreased risk for ESCC (OR=0.62, 95% CI =0.46-0.84 for ≤ 8 years ownership versus none).
Table 4.
Socioeconomic status and risk for upper gastrointestinal cancer
| ESCC | GCA | GNCA | ||||||
|---|---|---|---|---|---|---|---|---|
| Control (%) | Case (%) | OR (95% CI)# | Case (%) | OR (95% CI)# | Case (%) | OR (95% CI)# | ||
| Occupation | ||||||||
| Farmer | 760(50) | 306(51) | 1 | 275(46) | 1 | 141(45) | 1 | |
| Industrial worker | 390(26) | 137(23) | 0.94(0.59-1.22) | 142(24) | 0.98(0.76-1.26) | 66(21) | 1.04(0.74-1.48) | |
| Office worker | 239(16) | 108(18) | 1.29(0.97-1.72) | 147(24) | 1.54(1.18-2.02) | 83(26) | 2.30(1.63-3.24) | |
| Other | 125(8) | 48(8) | 0.85(0.59-1.22) | 35(6) | 0.88(0.58-1.32) | 26(8) | 1.20(0.74-1.93) | |
| Education | ||||||||
| 0-5 years | 650(47) | 228(41) | 1 | 228(41) | 1 | 97(33) | 1 | |
| 6-9 years | 561(40) | 244(44) | 0.75(0.60-0.94) | 202(36) | 0.90(0.71-1.13) | 133(46) | 0.69(0.52-0.93) | |
| ≥ 10 years | 174(13) | 83(15) | 1.12(0.82-1.51) | 128(23) | 1.98(1.49-2.64) | 61(21) | 1.50(1.05-2.13) | |
| Household size | ||||||||
| 0-3 | 806(53) | 284(47) | 1 | 292(49) | 1 | 145(46) | 1 | |
| >=4 | 708(47) | 316(53) | 1.29(1.06-1.58) | 307(51) | 1.31(1.07-1.59) | 171(54) | 1.20(0.93-1.55) | |
| Household income (Yuan/month) | ||||||||
| 0-259 | 378(25) | 159(26) | 1 | 145(24) | 1 | 61(19) | 1 | |
| 260-499 | 288(19) | 107(18) | 0.90(0.67-1.20) | 116(19) | 1.11(0.82-1.48) | 71(22) | 1.42(0.97-2.08) | |
| 500-999 | 429(28) | 164(27) | 0.92(0.70-1.20) | 155(26) | 0.99(0.75-1.31) | 84(26) | 1.16(0.80-1.69) | |
| >=1000 | 419(28) | 170(28) | 0.99(0.75-1.30) | 183(30) | 1.21(0.92-1.60) | 100(32) | 1.42(0.98-2.06) | |
| Own refrig years | ||||||||
| 0 | 948(63) | 397(66) | 1 | 370(62) | 1 | 193(61) | 1 | |
| 1-7 years | 276(18) | 121(20) | 1.03(0.80-1.33) | 107(18) | 1.00(0.77-1.31) | 58(18) | 1.06(0.75-1.48) | |
| ≥ 8 years | 289(19) | 82(14) | 0.62(0.46-0.84) | 122(20) | 1.13(0.86-1.49) | 65(20) | 1.14(0.80-1.62) | |
Adjusted for age (continuous), gender, geographic region (5 classes); OR: odds ratio; CI: confidence interval
ESCC: esophageal squamous cell carcinoma
GCA: gastric cardia adenocarcinoma
GNCA: gastric noncardia adenocarcinoma
Dietary factors
Intake frequencies of primary staple foods (not mutually exclusive) and meat and legumes were evaluated for the time periods both prior to and after 1984 (Table 5A). The primary staple cereal foods in controls shifted from corn (86% reported corn as a primary staple food) and sorghum (44%) before 1984, to wheat (98%) and rice (37%) after 1984. Corn as a primary staple food before 1984 was associated with lower GNCA risk (OR= 0.68, 95% CI= 0.42-0.86); while rice as a primary staple after 1984 was associated with decreased GCA risk (OR=0.78, 95% CI=0.63-0.98). Further adjustment for SES, smoking, and alcohol drinking did not alter these findings (data not shown).
Table 5A.
Dietary factors and risk for upper gastrointestinal cancer Cereal and meat and legumes intake and upper gastrointestinal cancer risk
| ESCC | GCA | GNCA | ||||||
|---|---|---|---|---|---|---|---|---|
| Control (%) | Case (%) | OR (95% CI)# | Case (%) | OR (95% CI)# | Case (%) | OR (95% CI)# | ||
| Before 1984 | ||||||||
| Primary staple food a | ||||||||
| Corn | Yes | 1305(86) | 511(85) | 0.93(0.68-1.27) | 514(86) | 1.03(0.75-1.42) | 239(76) | 0.60(0.42-0.86) |
| Millet | Yes | 196(13) | 78(13) | 1.06(0.77-1.47) | 99(16) | 1.32(0.97-1.79) | 44(14) | 1.06(0.71-1.60) |
| Wheat | Yes | 166(11) | 65(11) | 1.01(0.70-1.44) | 83(14) | 1.36(0.96-1.91) | 62(20) | 1.46(0.97-2.19) |
| Rice | Yes | 45(3) | 17(3) | 0.88(0..47-1.63) | 16(3) | 0.80(0.42-1.52) | 19(6) | 1.11(0.59-2.11) |
| Sorghum | Yes | 665(44) | 268(45) | 0.95(0.75-1.21) | 242(40) | 1.01(0.79-1.28) | 120(38) | 0.88(0.64-1.21) |
| Other | Yes | 21(1) | 6(1) | -- | 5(1) | - | 12(4) | - |
| Meat and legumes b | ||||||||
| Red meat | Any | 1447(96) | 568(95) | 0.82(0.48-1.38) | 578(96) | 0.84(0.47-1.52) | 297(94) | 0.73(0.39-1.37) |
| Chicken | Any | 1255(83) | 490(82) | 1.23(0.81-1.84) | 510(85) | 1.05(0.69-1.59) | 251(79) | 0.90(0.54-1.52) |
| Fish | Any | 1198(79) | 458(76) | 0.76(0.53-1.10) | 488(81) | 1.08(0.74-1.57) | 239(76) | 0.93(0.58-1.48) |
| Eggs | Any | 1460(96) | 586(98) | 1.75(0.92-3.31) | 588(98) | 1.82(0.91-3.63) | 305(97) | 1.22(0.60-2.50) |
| Beans or soy products |
Weekly vs. less than weekly |
478(32) | 182(30) | 0.94(0.78-1.17) | 201(33) | 1.06(0.86-1.31) | 104(33) | 1.08(0.83-1.41) |
| After 1984 | ||||||||
| Primary staple food c | ||||||||
| Corn | Yes | 66(4) | 16(3) | 0.61(0.34-1.09) | 31(5) | 0.99(0.62-1.59) | 12(4) | 0.94(0.48-1.82) |
| Millet | Yes | 80(5) | 32(5) | 0.99(0.64-1.54) | 49(8) | 1.42(0.96-2.09) | 19(6) | 1.20(0.69-2.06) |
| Wheat | Yes | 1489(98) | 595(99) | 1.74(0.65-4.71) | 584(98) | 0.74(0.37-1.49) | 308(97) | 0.72(0.31-1.67) |
| Rice | Yes | 566(37) | 206(34) | 0.82(0.66-1.02) | 185(31) | 0.78(0.63-0.98) | 130(41) | 1.19(0.90-1.57) |
| Sorghum | Yes | 5(0.3) | 2(0.3) | 0.99(0.18-5.42) | 5(1) | 2.01(0.53-7.62) | 0 | - |
| Other | Yes | 7(0.5) | 0 | - | 6(1) | - | 3(1) | - |
| Meat and legumes d | ||||||||
| Red meat | ||||||||
| Monthly/seldom/never | 598(40) | 231(38) | 1 | 204(34) | 1 | 101(32) | 1 | |
| Weekly | 513(34) | 203(34) | 1.14(0.89-1.46) | 214(36) | 1.21(0.95-1.55) | 126(40) | 1.62(1.18-2.24) | |
| >Weekly | 403(27) | 166(28) | 1.37(1.03-1.82) | 181(30) | 1.54(1.15-2.07) | 89(28) | 1.77(1.21-2.58) | |
| P for trend | 0.04 | <0.01 | 0.03 | |||||
| Chicken | ||||||||
| Never | 149(10) | 61(10) | 1 | 50(8) | 1 | 32(10) | 1 | |
| Seldom | 1140(75) | 476(79) | 1.37(1.03-1.82) | 468(78) | 0.90(0.55-1.47) | 246(78) | 0.92(0.50-1.70) | |
| Monthly | 110(7) | 33(6) | 1.14(0.89-1.46) | 36(6) | 1.004(0.52-1.93) | 20(6) | 0.74(0.33-1.68) | |
| Daily/weekly | 114(8) | 30(5) | 0.96(0.50-1.88) | 45(8) | 0.98(0.52-1.86) | 18(6) | 0.61(0.26-1.42) | |
| P for trend | 0.32 | 0.72 | 0.18 | |||||
| Fish | ||||||||
| Never | 179(12) | 91(15) | 1 | 59(10) | 1 | 37(12) | 1 | |
| Seldom | 1056(70) | 422(70) | 0.59(0.39-0.88) | 458(76) | 1.29(0.82-2.03) | 226(72) | 0.96(0.54-1.70) | |
| Monthly | 121(8) | 40(8) | 0.51(0.29-0.91) | 25(4) | 0.53(0.28-1.03) | 24(8) | 0.90(0.42-1.93) | |
| Daily/weekly | 157(10) | 47(8) | 0.50(0.28-0.88) | 57(10) | 0.86(0.48-1.55) | 29(9) | 0.93(0.45-1.95) | |
| P for trend | 0.06 | 0.06 | 0.83 | |||||
| Eggs | ||||||||
| Monthly/seldom/never | 274(18) | 105(18) | 1 | 98(16) | 1 | 63(20) | 1 | |
| Weekly | 429(28) | 170(28) | 1.04(0.77-1.41) | 180(30) | 1.10(0.81-1.49) | 101(32) | 0.93(0.64-1.34) | |
| Daily | 811(54) | 325(54) | 1.12(0.84-1.49) | 321(54) | 0.95(0.71-1.28) | 152(48) | 0.82(0.57-1.17) | |
| P for trend | 0.33 | 0.53 | 0.27 | |||||
| Beans or soy products | ||||||||
| Monthly/seldom/never | 354(23) | 138(23) | 1 | 123(20) | 1 | 78(25) | 1 | |
| Weekly | 754(50) | 319(53) | 1.10(0.86-1.42) | 323(54) | 1.16(0.89-1.50) | 178(56) | 1.02(0.74-1.40) | |
| Daily | 406(27) | 143(24) | 0.88(0.65-1.20) | 153(26) | 0.96(0.70-1.30) | 60(19) | 0.63(0.42-0.95) | |
| P for trend | 0.33 | 0.70 | 0.02 | |||||
Adjusted for age (continuous), gender, geographic region (5 classes); OR: odds ratio; CI: confidence interval
Primary staple food before 1984 were coded as yes or no consumption with not consumption as reference group. Primary staple foods were not mutual exclusive; 58% controls had more than one staple food before 1984. All these staple foods were adjust for each other
Red meat (pork, beef and lamb), chicken, fish, and eggs consumption before 1984 were coded as ever or never consumption with never as reference group; Red meat (pork, beef and lamb), chicken, fish, eggs, and soy products consumption before 1984 were adjusted for each other
Primary staple food after 1984 were coded as yes or no consumption with not consumption as reference group. Primary staple foods were not mutual exclusive; 46% controls had more than one staple food after 1984. All these staple foods were adjust for each other
Red meat (pork, beef and lamb), chicken, fish, eggs, and soy products consumption after 1984 were adjusted for each other
ESCC: esophageal squamous cell carcinoma
GCA: gastric cardia adenocarcinoma
GNCA: gastric noncardia adenocarcinoma
Intake of meat and legumes was rare before 1984: among controls, the median frequency of red meat consumption per year was 3; the median category of frequency of chicken, fish, eggs, and beans/soy consumption was “seldom”. After 1984, consumption of red meat increased dramatically (median frequency of 54 times per year), as did eggs (median category of frequency “daily”) and beans/soy (median category of frequency “weekly”). However, the median category of consumption remained unchanged for chicken and fish (“seldom”) (data not shown). Intake frequencies of these food items before 1984 were not associated with UGI cancer risk (Table 5A). After 1984, frequent intake of red meat (pork, beef, and lamb) was associated with increased risk of all three cancers. The most frequent eaters of fish (daily) had a 50% lower risk of ESCC, while persons who ate beans or soy most often (daily) had a 37% decreased risk of GNCA.
Consumption of fresh vegetables (median among controls, 1000 times/year) was more frequent than fruit (median among controls, 65 times/year) in our study population (Table 5B). Compared with the lowest intake quartiles, higher intakes of fresh vegetables and fruits were associated with monotonically decreased risk for all three cancers. Adjustment for potential confounders did not alter the findings. In contrast, persons who reported ever using milk or dairy products had 56% to 108% higher risk for UGI cancers than never users.
Table 5B.
Dietary factors and risk for upper gastrointestinal cancer Dairy products, fruits and vegetable intake and upper gastrointestinal cancer risk
| ESCC | GCA | GNCA | ||||||
|---|---|---|---|---|---|---|---|---|
| Control (%) | Case (%) | OR (95% CI)# | Case (%) | OR (95% CI)# | Case (%) | OR (95% CI)# | ||
|
Milk or dairy
products | ||||||||
| Ever vs. never | 901(60) | 415(69) | 1.56(1.27-1.92) | 455(76) | 2.08(1.67-2.59) | 222(70) | 1.69(1.29-2.21) | |
|
Fresh vegetables b (times/year) |
||||||||
| <774 | 379(25) | 194(32) | 1 | 201(34) | 1 | 107(34) | 1 | |
| 774-999 | 377(25) | 181(30) | 0.92(0.71-1.18) | 174(29) | 0.88(0.69-1.14) | 100(32) | 0.91(0.66-1.24) | |
| 1000-1229 | 377(25) | 130(22) | 0.64(0.49-0.83) | 122(20) | 0.62(0.47-0.81) | 77(24) | 0.70(0.51-0.98) | |
| >=1230 | 381(25) | 95(16) | 0.46(0.35-0.62) | 102(17) | 0.52(0.39-0.69) | 32(10) | 0.30(0.19-0.45) | |
| P for trend | <0.0001 | 0.0001 | 0.0001 | |||||
|
Fresh fruits c (times/year) |
||||||||
| <17 | 375(25) | 179(30) | 1 | 199(33) | 1 | 116(37) | 1 | |
| 17-64 | 382(25) | 177(30) | 0.92(0.71-1.18) | 182(30) | 0.92(0.72-1.18) | 92(29) | 0.74(0.54-1.01) | |
| 65-188 | 366(24) | 132(22) | 0.72(0.55-0.94) | 112(19) | 0.59(0.45-0.78) | 66(21) | 0.55(0.39-0.77) | |
| >=189 | 391(26) | 112(19) | 0.53(0.40-0.71) | 106(18) | 0.54(0.40-0.71) | 42(13) | 0.32(0.22-0.47) | |
| P for trend | <0.0001 | 0.0001 | 0.0001 | |||||
Adjusted for age (continuous), gender, geographic region (5 classes); OR: odds ratio; CI: confidence interval
Vegetable : potato, sweet potato, carrot, turnip, onion, bean pod, cabbage, cucumber, garlic bolt, gourds, squash, spinach, celery, rape, chives, green onion, tomato, eggplant; consumption was calculated as times per year
Fruit: apple, pear, grape, peach, apricot, watermelon, persimmon, hawthorn, orange, banana; consumption was calculated as times per year
ESCC: esophageal squamous cell carcinoma
GCA: gastric cardia adenocarcinoma
GNCA: gastric noncardia adenocarcinoma
Table 5C shows the intake frequencies of specific dietary factors in Shanxi Province. Intake of scalding hot food was associated with a monotonically increased risk for all three cancers. Daily consumption of scalding hot food was reported by 25% of controls and was a particularly strong risk factor for ESCC (OR= 2.74, 95% CI= 2.06-3.65), GCA (OR= 2.50, 95% CI=1.88-3.32), and GCNA (OR= 3.19. 95% CI =2.10-4.84). Moldy food intake was also associated with a greater than 50% increased risk for all three cancers. Consumptions of salted meat and pickled vegetables and pickled vegetable juice were associated with 20% to 40% increased risks for GCA and GNCA. Vinegar and seasoning use were both associated with monotonically decreased risks for all three cancers.
Table 5C.
Dietary factors and risk for upper gastrointestinal cancer Special food intake habits and upper gastrointestinal cancer risk
| ESCC | GCA | GNCA | ||||||
|---|---|---|---|---|---|---|---|---|
| Control (%) | Case (%) | OR (95% CI)# | Case (%) | OR (95% CI)# | Case (%) | OR (95% CI)# | ||
| Scalding hot food | ||||||||
| Never | 369(24) | 87(14) | 1 | 93(16) | 1 | 33(10) | ||
| Monthly/seldom | 647(43) | 228(38) | 1.46(1.11-1.94) | 218(36) | 1.39(1.05-1.83) | 152(48) | 2.60(1.74-3.87) | |
| Weekly | 125(8) | 43(7) | 1.46(0.96-2.23) | 58(10) | 1.92(1.30-2.84) | 22(7) | 1.90(1.07-3.40) | |
| Daily | 373(25) | 242(40) | 2.74(2.06-3.65) | 230(38) | 2.50(1.88-3.32) | 109(34) | 3.19(2.10-4.84) | |
| P for trend | <0..0001 | <0..0001 | <0.0001 | |||||
| Moldy food | ||||||||
| Ever vs. never | 461(30) | 240(40) | 1.55(1.27-1.89) | 241(40) | 1.53(1.26-1.87) | 131(41) | 1.66(1.29-2.14) | |
| Salted meat | ||||||||
| Ever vs. never | 570(38) | 248(41) | 1.20(0.98-1.46) | 258(43) | 1.24(1.02-1.51) | 149(47) | 1.46(1.14-1.87) | |
| Pickled vegetable and pickled vegetable juice | ||||||||
| Ever vs. never | 797(53) | 322(54) | 1.03(0.85-1.25) | 351(59) | 1.27(1.04-1.54) | 190(60) | 1.44(1.12-1.85) | |
| Vinegar | ||||||||
| <Monthly | 249(16) | 121(20) | 1 | 118(20) | 1 | 83(26) | 1 | |
| Weekly | 116(8) | 56(9) | 0.98(0.66-1.45) | 49(8) | 0.90(0.60-1.34) | 33(10) | 0.88(0.55-1.40) | |
| Daily | 1149(76) | 423(70) | 0.75(0.59-0.96) | 432(72) | 0.78(0.61-0.997) | 200(63) | 0.54(0.40-0.72) | |
| P for trend | 0.0148 | 0.0406 | <0.0001 | |||||
| Seasonings a | ||||||||
| Never | 97(6) | 47(8) | 1 | 38(6) | 1 | 26(8) | 1 | |
| Monthly/seldom | 580(38) | 264(44) | 0.93(0.64-1.36) | 264(44) | 1.20(0.80-1.80) | 139(44) | 0.93(0.58-1.49) | |
| Weekly | 333(22) | 124(21) | 0.77(0.51-1.16) | 128(21) | 1.01(0.65-1.55) | 72(23) | 0.83(0.50-1.37) | |
| Daily | 504(33) | 165(28) | 0.68(0.46-1.01) | 169(28) | 0.87(0.57-1.32) | 79(25) | 0.61(0.37-1.01) | |
| P for trend | 0.0035 | 0.0189 | 0.0053 | |||||
Adjusted for age (continuous), gender, geographic region (5 classes); OR: odds ratio; CI: confidence interval
Seasonings: black pepper, spicy pepper, garlic, ginger, cinnamon
ESCC: esophageal squamous cell carcinoma
GCA: gastric cardia adenocarcinoma
GNCA: gastric noncardia adenocarcinoma
Discussion
We evaluated lifestyle factors and risk of three UGI cancers in this relatively large case-control study (significant risk factors are summarized in Table 6). Tobacco smoking and alcohol drinking conferred little or no increased risk for these UGI cancers, while several SES and dietary factors showed significant associations for one or more of the UGI cancers. The strongest and most consistent risk factors across all three UGI cancer sites were the increased risk in persons who consumed scalding hot foods and the decreased risk observed in those who ate vegetables and fruits most frequently.
Table 6.
Summary of lifestyle factors and upper gastrointestinal cancer risk
| ESCC | GCA | GNCA | |
|---|---|---|---|
| Smoking | |||
| Current smoker | o | − | o |
| Higher intensity | o | o | o |
| Longer duration | o | − | o |
| Higher pack-years | + | o | o |
| Alcohol | |||
| Beer | o | o | + |
| Socioeconomic status | |||
| Office worker | o | + | + |
| Higher education | + | + | + |
| Larger household size | + | + | + |
| Refrigerator usage | − | o | o |
| Diet | |||
| Scalding hot food | + | + | + |
| Moldy food | + | + | + |
| Salted meat | + | + | + |
| Pickled vegetables/ pickled vegetable juice | + | + | + |
| Vinegar | − | − | − |
| Seasonings | − | − | − |
| Milk and dairy products | + | + | + |
| Red meat >1984 | + | + | + |
| Fish >1984 | − | o | o |
| Beans and soy products >1984 | o | o | − |
| Fresh vegetables | − | − | − |
| Fresh fruits | − | − | − |
significantly increased risk
significantly decreased risk
no-significant association
ESCC: esophageal squamous cell carcinoma
GCA: gastric cardia adenocarcinoma
GNCA: gastric noncardia adenocarcinoma
Smoking is a dominant risk factor for ESCC in the West [3; 13; 21; 22; 23], and some parts of Asia [24; 25] . However, smoking exposures confer little risk for ESCC in the areas of the world where this disease occurs in epidemic proportions with rates in excess of 100 per 100,000 (ie, the Taihang mountain region of north central China which includes Shanxi and Linxian; northern Iran around the Caspian Littoral) [9; 16; 26]. Analytic epidemiologic studies from the West show that tobacco smoking is associated with a 3- to 7-fold increased risk of ESCC in current smokers [27]; in contrast, increased ESCC risk from smoking is minimal in north central China and Iran, with ORs of only ~1.5 [27]. In these high-risk areas, women have rates that are as high as men (ie, northern Iran) [28] or nearly as high as men (Taihang mountain region of China) [29], despite the fact that the women are virtually all non-smokers.
Consistent with previous studies of ESCC in high-risk regions [27], we found only limited evidence of an association between tobacco and risk, with a 50% elevated risk among men in the highest quartile of cumulative lifetime tobacco exposure. Further adjustment for potential confounders did not alter our findings. The most likely explanation for these results is that people in the high-risk region of Shanxi are exposed to carcinogens similar to those found in tobacco (ie, polycyclic aromatic hydrocarbons [PAHs], nitrosamines) through alternative environmental exposures, such as general outdoor air pollution or indoor air pollution resulting from cooking or heating practices, and that the effect of these alternative environmental exposures overwhelms any effect of tobacco.
Alcohol consumption is also a dominant risk factor for ESCC in the West where studies consistently show a 3 to 5-fold increased risk [13; 27]. Similar to tobacco, however, previous studies of alcohol in Linxian and Shanxi have also shown little or no association with risk for UGI cancers [9; 16; 30]. In the present study we found no association with ever use of alcohol for any UGI cancer site; our only notable finding was for beer consumption, where GNCA risk was consistently elevated by about 50% regardless of season or year of consumption. Confounding did not explain these findings either, as further adjustment for multiple covariates did not change these results. The apparent discrepancy for the role of alcohol in UGI cancers between the West and high-risk areas of the world may be due to the consumption rate, which is high in the West [31] but relatively low in rural areas of Asia as well as other high-risk areas that have religious proscription of alcohol, such as Iran. It is also possible that acetaldehyde, the putative carcinogen formed by the metabolism of alcohol, is still responsible for elevated risk in high-risk populations as a result of alternative non-alcoholic sources of exposure (eg, yogurt).
Previous studies have shown that workers with occupational exposure to toxic substances have higher risk of UGI cancers when compared to office workers [32; 33]. Similarly, lower education has also been associated with increased risk for these cancers [9; 34]. In the present study, however, we found the opposite: both office workers and persons with higher education were at increased risk of UGI cancer. Although these characteristics are correlated, mutual adjustment for both factors (as well as other potential confounders) showed that they were independent risk factors. The reason our results differ from earlier reports is not clear and may be due to unknown or unmeasured confounding factors (eg, physical activity [35]), or the role of chance; another possibility is that our use of neighborhood controls may have resulted in over matching for certain factors (such as SES) and reduced our ability to observe differences for these factors. Larger family sizes have been associated with increased esophageal and gastric cancer risk in several previous studies [36; 37; 38]. We also observed increased UGI cancer risk among individuals with larger families, which may potentially be due to an infectious etiology enhanced by a crowded environment [39]. Studies have suggested that Helicobacter pylori (Hp), Epstein-Barr virus, human papilloma virus, and some fungi may be related to UGI cancers [40; 41; 42]; however, at this time only Hp is an established human carcinogen for gastric cancer [40]. Unfortunately, we do not have specific information on infections in the subjects in the current study.
China experienced an economic reformation from the late 1970s to the early 1980s, which resulted in dramatic economic development. The impact of this reformation was relatively rapid, over a matter of a few years, and improved dietary intake was one important benefit of this development. As a result, we observed major shifts in the primary staple foods from corn and sorghum before 1984 to wheat and rice after; moreover, the frequencies of meat and legumes also increased greatly after 1984. Though hard, scratchy, and starchy staple foods, like maize and sorghum, have reportedly been associated with increased esophageal cancer risk due to the physical irritation and/or nutritional deficiencies [43; 44; 45], we found an association opposite: corn as primary staple food before 1984 was associated with decreased GNCA risk. In addition, persons who reported rice as a primary staple after 1984 had lower risk of GCA. Neither of these results related to primary staple grains was significant or consistent across all three cancer sites, suggesting that further study is needed to sort out these interesting but inconsistent findings.
Parallel to the shift in staple foods, consumption of animal protein was considerably improved after 1984. Numerous studies have reported positive associations between red meat intake and UGI cancers [46; 47; 48], and we also observed consistently elevated risk for all three UGI cancers among red meat consumers after 1984, including after adjustment for SES, smoking, alcohol, and other potential confounding factors. The most frequently suggested mechanistic explanation for this association is that carcinogenic compounds were generated from high temperature cooking (eg, N-nitroso compounds, heterocyclic amines, and PAHs), as well as free radicals from heme iron [49]. Before 1984, the intake of meat was as low as that observed in Linxian during the same time period [9], and no associations with risk of UGI cancer were observed there either.
Though not consistent, some studies have reported positive associations between intake of dairy products and gastric cancer risk [50; 51]. Intake of milk or dairy products was associated with increased risk of all three UGI cancers studied here. Adjustment for other covariates did not diminish these effects. Possible explanation for these associations may relate to inflammation due to lactose intolerance, or carcinogen contamination of the milk products [52].
Fresh vegetable and fruit consumption are well-established protective factors for UGI cancer [53]. As expected, cancer risk decreased monotonically with increased intake of fruits and vegetables across all three of the UGI cancer sites studied here. Modulation of xenobiotic-metabolizing genes and anti-oxidation from the vitamins and phytochemicals in fruits and vegetables are often cited as potential mechanisms [54].
Consumption of hot foods and beverages has consistently been associated with increased risk for esophageal cancer (reviewed by Islami et al) [55]. We observed markedly increased risk for all three UGI cancers studied here, which adds evidence to the thermal damage etiologic hypothesis. Importantly, daily consumption of scalding hot food was reported by 25% of this population, thus making this practice a target for risk reduction by education and behavior modification.
Other special foods and food practices, like consumption of moldy food, salted meat, pickled vegetables, and pickled vegetable juice, have also been associated with elevated risk for UGI cancers in previous reports [30; 56; 57; 58; 59]. We observed similar positive associations between consumption of these foods and risk for UGI cancers. Contamination by N-nitroso compounds and mycotoxins are the most frequently proposed link between these food items and increased risk of UGI cancers [60]. In contrast, we observed protective effects in persons who most frequently consumed vinegar and seasonings typical of Chinese cuisine (eg, black pepper, spicy pepper, garlic, ginger, cinnamon); these findings are consistent with previous reports [45; 58; 61; 62; 63]. The active phytochemicals in these foods may underline their presumptive cancer preventive action.
There are several notable advantages to our study: we had a relatively large sample size with exclusively histologically-confirmed cancer cases, participation rates for both cases and controls were very high, and we employed neighborhood-matched controls. These advantages allowed us to obtain unbiased estimates of modest effects not captured by small studies. The detailed frequency and duration information on cigarettes and pipe smoking allowed us to estimate cumulative lifetime smoking exposure. The dietary patterns captured for both before and after 1984 provided better estimates for the effects of these factors. Our study collected information on specific dietary elements consumed in Shanxi, which provided an opportunity to evaluate important lifestyle factors specific to this population. In addition, we were able to simultaneously evaluate relations with cancers at three adjacent but anatomically different sites using identical study methods.
Some limitations to our study are also evident: susceptibility to recall bias is inherent in the case-control approach; the use of neighborhood controls may have resulted in over matching for selected risk factors; we did not collect information on all known potential confounders (eg, physical activity); our dietary questionnaire was not validated, and information was not collected on portion size or cooking practices, which may have resulted in residual confounding; and we recruited only cases who had surgery as their initial and primary therapy, so our results may not be generalizable to all UGI cancer cases.
Conclusions
This large case-control study of UGI cancers corroborated prior results that showed that tobacco and alcohol have only minor roles in these cancers in north central China, where rates of esophageal and gastric cancer are among the highest in the world. The study also confirmed the protective role for vegetable and fruit consumption, and highlighted the importance of thermal damage as a strong risk factor for UGI cancer.
Acknowledgements
This study was supported by the Intramural Research Program of National Institutes of Health, National Cancer Institute, Division of Cancer Epidemiology and Genetics. The funding unit had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
We thank all the subjects who participated in the study and all the collaborators who contributed to its success.
We thank the editorial assistance of the Fellows Editorial Board of National Institutes of Health.
Footnotes
Conflict of interest The authors declare that they have no conflict of interest.
Author contributions NH, X-YH, CG, TD, AMG, and PRT designed and conducted the study; YG conducted the statistical analyses; AMG and PRT oversaw the statistical analyses; YG wrote the first draft of the manuscript; all authors contributed the writing and/or editing of the final manuscript.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- [1].Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLoS Med. 2006;3:e442. doi: 10.1371/journal.pmed.0030442. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Qiao YL, Hou J, Yang L, He YT, Liu YY, Li LD, Li SS, Lian SY, Dong ZW. The trends and preventive strategies of esophageal cancer in high-risk areas of Taihang Mountains, China. Zhongguo Yi Xue Ke Xue Yuan Xue Bao. 2001;23:10–14. [PubMed] [Google Scholar]
- [3].Blot WJ, McLaughlin JK. The changing epidemiology of esophageal cancer. Semin Oncol. 1999;26:2–8. [PubMed] [Google Scholar]
- [4].Brown LM, Devesa SS. Epidemiologic trends in esophageal and gastric cancer in the United States. Surg Oncol Clin N Am. 2002;11:235–256. doi: 10.1016/s1055-3207(02)00002-9. [DOI] [PubMed] [Google Scholar]
- [5].Crew KD, Neugut AI. Epidemiology of upper gastrointestinal malignancies. Semin Oncol. 2004;31:450–464. doi: 10.1053/j.seminoncol.2004.04.021. [DOI] [PubMed] [Google Scholar]
- [6].Devesa SS, Blot WJ, Fraumeni JF., Jr. Changing patterns in the incidence of esophageal and gastric carcinoma in the United States. Cancer. 1998;83:2049–2053. [PubMed] [Google Scholar]
- [7].Wu AH, Wan P, Bernstein L. A multiethnic population-based study of smoking, alcohol and body size and risk of adenocarcinomas of the stomach and esophagus (United States) Cancer Causes Control. 2001;12:721–732. doi: 10.1023/a:1011290704728. [DOI] [PubMed] [Google Scholar]
- [8].Siman JH, Forsgren A, Berglund G, Floren CH. Tobacco smoking increases the risk for gastric adenocarcinoma among Helicobacter pylori-infected individuals. Scand J Gastroenterol. 2001;36:208–213. doi: 10.1080/003655201750065988. [DOI] [PubMed] [Google Scholar]
- [9].Tran GD, Sun XD, Abnet CC, Fan JH, Dawsey SM, Dong ZW, Mark SD, Qiao YL, Taylor PR. Prospective study of risk factors for esophageal and gastric cancers in the Linxian general population trial cohort in China. Int J Cancer. 2005;113:456–463. doi: 10.1002/ijc.20616. [DOI] [PubMed] [Google Scholar]
- [10].Lee CH, Lee JM, Wu DC, Hsu HK, Kao EL, Huang HL, Wang TN, Huang MC, Wu MT. Independent and combined effects of alcohol intake, tobacco smoking and betel quid chewing on the risk of esophageal cancer in Taiwan. Int J Cancer. 2005;113:475–482. doi: 10.1002/ijc.20619. [DOI] [PubMed] [Google Scholar]
- [11].Nishino Y, Inoue M, Tsuji I, Wakai K, Nagata C, Mizoue T, Tanaka K, Tsugane S. Tobacco smoking and gastric cancer risk: an evaluation based on a systematic review of epidemiologic evidence among the Japanese population. Jpn J Clin Oncol. 2006;36:800–807. doi: 10.1093/jjco/hyl112. [DOI] [PubMed] [Google Scholar]
- [12].Tredaniel J, Boffetta P, Buiatti E, Saracci R, Hirsch A. Tobacco smoking and gastric cancer: review and meta-analysis. Int J Cancer. 1997;72:565–573. doi: 10.1002/(sici)1097-0215(19970807)72:4<565::aid-ijc3>3.0.co;2-o. [DOI] [PubMed] [Google Scholar]
- [13].Enzinger PC, Mayer RJ. Esophageal cancer. N Engl J Med. 2003;349:2241–2252. doi: 10.1056/NEJMra035010. [DOI] [PubMed] [Google Scholar]
- [14].Hashibe M, Boffetta P, Janout V, Zaridze D, Shangina O, Mates D, Szeszenia-Dabrowska N, Bencko V, Brennan P. Esophageal cancer in Central and Eastern Europe: tobacco and alcohol. Int J Cancer. 2007;120:1518–1522. doi: 10.1002/ijc.22507. [DOI] [PubMed] [Google Scholar]
- [15].Zambon P, Talamini R, Vecchia C. La, Maso L. Dal, Negri E, Tognazzo S, Simonato L, Franceschi S. Smoking, type of alcoholic beverage and squamous-cell oesophageal cancer in northern Italy. Int J Cancer. 2000;86:144–149. doi: 10.1002/(sici)1097-0215(20000401)86:1<144::aid-ijc23>3.0.co;2-b. [DOI] [PubMed] [Google Scholar]
- [16].Wang YP, Han XY, Su W, Wang YL, Zhu YW, Sasaba T, Nakachi K, Hoshiyama Y, Tagashira Y. Esophageal cancer in Shanxi Province, People’s Republic of China: a case-control study in high and moderate risk areas. Cancer Causes Control. 1992;3:107–113. doi: 10.1007/BF00051650. [DOI] [PubMed] [Google Scholar]
- [17].Mark SD, Qiao YL, Dawsey SM, Wu YP, Katki H, Gunter EW, Fraumeni JF, Jr., Blot WJ, Dong ZW, Taylor PR. Prospective study of serum selenium levels and incident esophageal and gastric cancers. Journal of the National Cancer Institute. 2000;92:1753–1763. doi: 10.1093/jnci/92.21.1753. [DOI] [PubMed] [Google Scholar]
- [18].Taylor PR, Qiao YL, Abnet CC, Dawsey SM, Yang CS, Gunter EW, Wang W, Blot WJ, Dong ZW, Mark SD. Prospective study of serum vitamin E levels and esophageal and gastric cancers. Journal of the National Cancer Institute. 2003;95:1414–1416. doi: 10.1093/jnci/djg044. [DOI] [PubMed] [Google Scholar]
- [19].Gao Y, Hu N, Han X, Giffen C, Ding T, Goldstein A, Taylor P. Family history of cancer and risk for esophageal and gastric cancer in Shanxi, China. BMC Cancer. 2009;9:269. doi: 10.1186/1471-2407-9-269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Gao Y, Hu N, Han X, Giffen C, Ding T, Goldstein AM, Taylor PR. Jasmine tea consumption and upper gastrointestinal cancer in China. Cancer Causes Control. 2009 doi: 10.1007/s10552-009-9394-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Blot WJ. Esophageal cancer trends and risk factors. Semin Oncol. 1994;21:403–410. [PubMed] [Google Scholar]
- [22].Kollarova H, Machova L, Horakova D, Janoutova G, Janout V. Epidemiology of esophageal cancer--an overview article. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2007;151:17–20. doi: 10.5507/bp.2007.003. [DOI] [PubMed] [Google Scholar]
- [23].Pelucchi C, Gallus S, Garavello W, Bosetti C, Vecchia C. La. Alcohol and tobacco use, and cancer risk for upper aerodigestive tract and liver. Eur J Cancer Prev. 2008;17:340–344. doi: 10.1097/CEJ.0b013e3282f75e91. [DOI] [PubMed] [Google Scholar]
- [24].Gao YT, McLaughlin JK, Blot WJ, Ji BT, Benichou J, Dai Q, Fraumeni JF., Jr. Risk factors for esophageal cancer in Shanghai, China. I. Role of cigarette smoking and alcohol drinking. Int J Cancer. 1994;58:192–196. doi: 10.1002/ijc.2910580208. [DOI] [PubMed] [Google Scholar]
- [25].Jiang JM, Zeng XJ, Chen JS, Li JY, Zhang KL, Wu YP, Liu BQ. Smoking and mortality from esophageal cancer in China: a large case-control study of 19,734 male esophageal cancer deaths and 104,846 living spouse controls. Int J Cancer. 2006;119:1427–1432. doi: 10.1002/ijc.21887. [DOI] [PubMed] [Google Scholar]
- [26].Nasrollahzadeh D, Kamangar F, Aghcheli K, Sotoudeh M, Islami F, Abnet CC, Shakeri R, Pourshams A, Marjani HA, Nouraie M, Khatibian M, Semnani S, Ye W, Boffetta P, Dawsey SM, Malekzadeh R. Opium, tobacco, and alcohol use in relation to oesophageal squamous cell carcinoma in a high-risk area of Iran. British journal of cancer. 2008;98:1857–1863. doi: 10.1038/sj.bjc.6604369. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Kamangar F, Chow WH, Abnet CC, Dawsey SM. Environmental causes of esophageal cancer. Gastroenterol Clin North Am. 2009;38:27–57. vii. doi: 10.1016/j.gtc.2009.01.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Islami F, Kamangar F, Aghcheli K, Fahimi S, Semnani S, Taghavi N, Marjani HA, Merat S, Nasseri-Moghaddam S, Pourshams A, Nouraie M, Khatibian M, Abedi B, Brazandeh MH, Ghaziani R, Sotoudeh M, Dawsey SM, Abnet CC, Taylor PR, Malekzadeh R. Epidemiologic features of upper gastrointestinal tract cancers in Northeastern Iran. British journal of cancer. 2004;90:1402–1406. doi: 10.1038/sj.bjc.6601737. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Yang CS. Research on esophageal cancer in China: a review. Cancer research. 1980;40:2633–2644. [PubMed] [Google Scholar]
- [30].Guo W, Blot WJ, Li JY, Taylor PR, Liu BQ, Wang W, Wu YP, Zheng W, Dawsey SM, Li B, et al. A nested case-control study of oesophageal and stomach cancers in the Linxian nutrition intervention trial. Int J Epidemiol. 1994;23:444–450. doi: 10.1093/ije/23.3.444. [DOI] [PubMed] [Google Scholar]
- [31].Cochranem H.C. Johanne. Katherine Conigrave, Wei Hao, Alcohol use in China. Alcohol & alcoholism. 2003;38:537–542. doi: 10.1093/alcalc/agg111. [DOI] [PubMed] [Google Scholar]
- [32].Bosetti C, Franceschi S, Negri E, Talamini R, Tomei F, Vecchia C. La. Changing socioeconomic correlates for cancers of the upper digestive tract. Ann Oncol. 2001;12:327–330. doi: 10.1023/a:1011180524985. [DOI] [PubMed] [Google Scholar]
- [33].Parent ME, Siemiatycki J, Fritschi L. Occupational exposures and gastric cancer. Epidemiology. 1998;9:48–55. [PubMed] [Google Scholar]
- [34].Islami F, Kamangar F, Nasrollahzadeh D, Aghcheli K, Sotoudeh M, Abedi-Ardekani B, Merat S, Nasseri-Moghaddam S, Semnani S, Sepehr A, Wakefield J, Moller H, Abnet CC, Dawsey SM, Boffetta P, Malekzadeh R. Socio-economic status and oesophageal cancer: results from a population-based case-control study in a high-risk area. Int J Epidemiol. 2009 doi: 10.1093/ije/dyp195. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].Sjodahl K, Jia C, Vatten L, Nilsen T, Hveem K, Lagergren J. Body mass and physical activity and risk of gastric cancer in a population-based cohort study in Norway. Cancer Epidemiol Biomarkers Prev. 2008;17:135–140. doi: 10.1158/1055-9965.EPI-07-0704. [DOI] [PubMed] [Google Scholar]
- [36].Altieri A, Hemminki K. Number of siblings and the risk of solid tumours: a nation-wide study. British journal of cancer. 2007;96:1755–1759. doi: 10.1038/sj.bjc.6603760. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [37].Shi R, Xu S, Zhang H, Ding Y, Sun G, Huang X, Chen X, Li X, Yan Z, Zhang G. Prevalence and risk factors for Helicobacter pylori infection in Chinese populations. Helicobacter. 2008;13:157–165. doi: 10.1111/j.1523-5378.2008.00586.x. [DOI] [PubMed] [Google Scholar]
- [38].Wei WQ, Abnet CC, Lu N, Roth MJ, Wang GQ, Dye BA, Dong ZW, Taylor PR, Albert P, Qiao YL, Dawsey SM. Risk factors for oesophageal squamous dysplasia in adult inhabitants of a high risk region of China. Gut. 2005;54:759–763. doi: 10.1136/gut.2004.062331. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [39].Altieri A, Castro F, Bermejo JL, Hemminki K. Association between number of siblings and nervous system tumors suggests an infectious etiology. Neurology. 2006;67:1979–1983. doi: 10.1212/01.wnl.0000247036.98444.38. [DOI] [PubMed] [Google Scholar]
- [40].Forman D. Helicobacter pylori and gastric cancer. Scand J Gastroenterol Suppl. 1996;220:23–26. [PubMed] [Google Scholar]
- [41].Mladenova I, Pellicano R. Infectious agents and gastric tumours. An increasing role for Epstein-Barr virus. Panminerva Med. 2003;45:183–188. [PubMed] [Google Scholar]
- [42].Ribeiro U, Jr., Posner MC, Safatle-Ribeiro AV, Reynolds JC. Risk factors for squamous cell carcinoma of the oesophagus. Br J Surg. 1996;83:1174–1185. [PubMed] [Google Scholar]
- [43].Ghadirian P, Ekoe JM, Thouez JP. Food habits and esophageal cancer: an overview. Cancer Detect Prev. 1992;16:163–168. [PubMed] [Google Scholar]
- [44].Mayne ST, Risch HA, Dubrow R, Chow WH, Gammon MD, Vaughan TL, Farrow DC, Schoenberg JB, Stanford JL, Ahsan H, West AB, Rotterdam H, Blot WJ, Fraumeni JF., Jr. Nutrient intake and risk of subtypes of esophageal and gastric cancer. Cancer Epidemiol Biomarkers Prev. 2001;10:1055–1062. [PubMed] [Google Scholar]
- [45].Xibib S, Meilan H, Moller H, Evans HS, Dixin D, Wenjie D, Jianbang L. Risk factors for oesophageal cancer in Linzhou, China: a case-control study. Asian Pac J Cancer Prev. 2003;4:119–124. [PubMed] [Google Scholar]
- [46].Cross AJ, Leitzmann MF, Gail MH, Hollenbeck AR, Schatzkin A, Sinha R. A prospective study of red and processed meat intake in relation to cancer risk. PLoS Med. 2007;4:e325. doi: 10.1371/journal.pmed.0040325. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [47].Gonzalez CA, Jakszyn P, Pera G, Agudo A, Bingham S, Palli D, Ferrari P, Boeing H, del Giudice G, Plebani M, Carneiro F, Nesi G, Berrino F, Sacerdote C, Tumino R, Panico S, Berglund G, Siman H, Nyren O, Hallmans G, Martinez C, Dorronsoro M, Barricarte A, Navarro C, Quiros JR, Allen N, Key TJ, Day NE, Linseisen J, Nagel G, Bergmann MM, Overvad K, Jensen MK, Tjonneland A, Olsen A, Bueno-de-Mesquita HB, Ocke M, Peeters PH, Numans ME, Clavel-Chapelon F, Boutron-Ruault MC, Trichopoulou A, Psaltopoulou T, Roukos D, Lund E, Hemon B, Kaaks R, Norat T, Riboli E. Meat intake and risk of stomach and esophageal adenocarcinoma within the European Prospective Investigation Into Cancer and Nutrition (EPIC) Journal of the National Cancer Institute. 2006;98:345–354. doi: 10.1093/jnci/djj071. [DOI] [PubMed] [Google Scholar]
- [48].Silvera S.A. Navarro, Mayne ST, Risch H, Gammon MD, Vaughan TL, Chow WH, Dubrow R, Schoenberg JB, Stanford JL, West AB, Rotterdam H, Blot WJ, Fraumeni JF., Jr. Food group intake and risk of subtypes of esophageal and gastric cancer. Int J Cancer. 2008;123:852–860. doi: 10.1002/ijc.23544. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [49].Lagiou P, Talamini R, Samoli E, Lagiou A, Ahrens W, Pohlabeln H, Benhamou S, Bouchardy C, Slamova A, Schejbalova M, Merletti F, Richiardi L, Kjaerheim K, Agudo A, Castellsague X, Macfarlane TV, Macfarlane GJ, Biggs AM, Barzan L, Canova C, Simonato L, Lowry RJ, Conway DI, McKinney PA, Znaor A, McCartan BE, Healy CM, Marron M, Hashibe M, Brennan P. Diet and upper-aerodigestive tract cancer in Europe: The ARCAGE study. Int J Cancer. 2008 doi: 10.1002/ijc.24246. [DOI] [PubMed] [Google Scholar]
- [50].Shrier I, Szilagyi A, Correa JA. Impact of lactose containing foods and the genetics of lactase on diseases: an analytical review of population data. Nutr Cancer. 2008;60:292–300. doi: 10.1080/01635580701745301. [DOI] [PubMed] [Google Scholar]
- [51].Ward MH, Lopez-Carrillo L. Dietary factors and the risk of gastric cancer in Mexico City. Am J Epidemiol. 1999;149:925–932. doi: 10.1093/oxfordjournals.aje.a009736. [DOI] [PubMed] [Google Scholar]
- [52].Alonso-Amelot ME. The link between bracken fern and stomach cancer: milk. Nutrition. 1997;13:694–696. doi: 10.1016/s0899-9007(97)83017-1. [DOI] [PubMed] [Google Scholar]
- [53].Food, nutrition, and the prevention of cancer: a global perspective. 2007.
- [54].Tsugane S, Sasazuki S. Diet and the risk of gastric cancer: review of epidemiological evidence. Gastric Cancer. 2007;10:75–83. doi: 10.1007/s10120-007-0420-0. [DOI] [PubMed] [Google Scholar]
- [55].Islami F, Boffetta P, Ren JS, Pedoeim L, Khatib D, Kamangar F. High-temperature beverages and foods and esophageal cancer risk--a systematic review. Int J Cancer. 2009;125:491–524. doi: 10.1002/ijc.24445. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [56].Cheng KK, Day NE, Duffy SW, Lam TH, Fok M, Wong J. Pickled vegetables in the aetiology of oesophageal cancer in Hong Kong Chinese. Lancet. 1992;339:1314–1318. doi: 10.1016/0140-6736(92)91960-g. [DOI] [PubMed] [Google Scholar]
- [57].Cheng KK, Duffy SW, Day NE, Lam TH. Oesophageal cancer in never-smokers and never-drinkers. Int J Cancer. 1995;60:820–822. doi: 10.1002/ijc.2910600616. [DOI] [PubMed] [Google Scholar]
- [58].Hung HC, Huang MC, Lee JM, Wu DC, Hsu HK, Wu MT. Association between diet and esophageal cancer in Taiwan. J Gastroenterol Hepatol. 2004;19:632–637. doi: 10.1111/j.1440-1746.2004.03346.x. [DOI] [PubMed] [Google Scholar]
- [59].You WC, Blot WJ, Chang YS, Ershow AG, Yang ZT, An Q, Henderson B, Xu GW, Fraumeni JF, Jr., Wang TG. Diet and high risk of stomach cancer in Shandong, China. Cancer research. 1988;48:3518–3523. [PubMed] [Google Scholar]
- [60].Eichholzer M, Gutzwiller F. Dietary nitrates, nitrites, and N-nitroso compounds and cancer risk: a review of the epidemiologic evidence. Nutr Rev. 1998;56:95–105. doi: 10.1111/j.1753-4887.1998.tb01721.x. [DOI] [PubMed] [Google Scholar]
- [61].Chen YK, Lee CH, Wu IC, Liu JS, Wu DC, Lee JM, Goan YG, Chou SH, Huang CT, Lee CY, Hung HC, Yang JF, Wu MT. Food intake and the occurrence of squamous cell carcinoma in different sections of the esophagus in Taiwanese men. Nutrition. 2009;25:753–761. doi: 10.1016/j.nut.2009.02.002. [DOI] [PubMed] [Google Scholar]
- [62].Kim JY, Kwon O. Garlic intake and cancer risk: an analysis using the Food and Drug Administration’s evidence-based review system for the scientific evaluation of health claims. Am J Clin Nutr. 2009;89:257–264. doi: 10.3945/ajcn.2008.26142. [DOI] [PubMed] [Google Scholar]
- [63].Pourfarzi F, Whelan A, Kaldor J, Malekzadeh R. The role of diet and other environmental factors in the causation of gastric cancer in Iran--a population based study. Int J Cancer. 2009;125:1953–1960. doi: 10.1002/ijc.24499. [DOI] [PubMed] [Google Scholar]