Correlation Of Abo Blood Groups, Body Mass Index (kg/m~2) And Esophageal Cancer

1 BACKGROUND AND OBJECTSEsophageal cancer(EC) is one of the sixth most common malignant diseases worldwide.The remarkable geographical distribution is the striking characteristic for esophageal cancer.Linzhou City(formerly Linxian County) and its neighboring counties in Henan Province,in northern China have been well known as the highest incidence and mortality area for EC in the world.From these high-incidence areas, extending outward on all sides,the figures gradually decrease,giving an impression of irregular concentric belt.The ratio for EC incidence between the high- and low-risk areas could be as high as 500:1.Another remarkable epidemiological characteristic for EC is the apparent familial aggregation.Recent studies by us demonstrate that familial aggregation in Linzhou is obviously higher than in other areas within Henan and in other provinces,and that the prevalence of EC and the prevalence of esophageal precancerous lesions in the migrated subjects is similar with Linzhou subjects over 100 years from Linzhou.The remarkable geographical distribution and familial aggregation,especially migrant studies suggest that both environmental and genetic factors may play important roles in esophageal carcinogenesis.As an important genetic phenotype,ABO blood group system receives more and more attention.The ABO blood group system was discovered by Karl Land Steiner in 1902. Since then,scientists have speculated on an association between the ABO blood group system and incidence and prognosis in cancer and other disease.However,the results are not consistent.The relationship between blood group and EC in high- and low-incidence, and the corresponding tumor family history is largely unknown.Recent studies indicate that ABO blood group is related with obesity.It has been generally accepted that obesity is one of the main risk factors for adenocarcinoma of the esophagus and gastric cardia.Obesity can promote gastroesophageal reflux disease by increasing intra-abdominal pressure.Gastroesophageal reflux predisposes to Barrett's esophagus,which is a metaplastic precursor state for adenocarcinoma at the esophageal-gastric junction.In western countries,the esophageal and gastric cardia adenocarcinoma has been increasing rapidly in the past 20 years.It is noteworthy that economic and living conditions have improved greatly in China in the past 20 years.With increased dietary fat consumption,aging,and prevention strategies,especially the Western food style introduced into China,gastroesophageal reflux disease has been reported to show increasing tendency,but the histopathological pattern for EC in China is stable,suggesting that there may be different mechanisms involved in Chinese subjects for carcinogenesis at the esophageal-gastric junction compared with western countries.The present study was undertaken to evaluate the association among the ABO blood group and body mass index(BMI) and the risk of EC to further elucidate the hereditary susceptibility of EC and to provide important clues and information in identifying the biomarkers for high-risk subject screening and early detection in EC.2 MATERIALS AND METHODS2.1 Subjects 2.1.1 EC patients:All the 10,440 EC patients were enrolled from Anyang,Linzhou, and Huixian in Henan,Cixia in Hebei,Yanting in Sichuan and Changzhi,Shanxi et al. Of the patients,4083 cases had detailed information of ABO blood type(2501 male, with a mean age of 57±9 years;1582 female,with a mean age of 57±9 years).1655 cases had detailed information of height and weight(1049 male,with a mean age of 60±9 years;606 female,with a mean age of 62±9 years).All the EC patients were confirmed with primary squamous cell carcinoma by histopathology.2.1.2 Healthy controls:Eleven hundred and thirty four healthy subjects from large scale of mass survey were enrolled as healthy control group.Of the patients,720 subjects have detailed information of ABO blood type(309 male,with a mean age of 51±10 years;411 female,with a mean age of 55±10 years).414 cases had detailed information of height and weight(185 male,with a mean age of 45±12 years;229 female,with a mean age of 47±12 years).2.2 Classification for high- and low- incidence areas for ECThere were five different areas involved in this study:Group 1:Taihang Mountain areas(the southern of Taihang mountain at the junction of Henan,Hebei and Shanxi provinces);Group 2:Other high incidence areas in three provinces(the high incidence areas in Henan,Hebei and Shanxi province,except Taihang areas);Group 3:Low incidence areas in three provinces(the low incidence areas for EC in Henan,Hebei and Shanxi provinces);Group 4:High incidence areas outside three provinces(the high incidence areas for EC in other province,outside the Henan,Hebei and Shanxi provinces)Group 5:Low incidence areas outside three provinces(the low incidence areas for EC in other province,outside the Henan,Hebei and Shanxi provinces)2.3 Epidemiological investigationThe epidemiologic survey questionnaire was used and EC patients or their family members were interviewed from 2005 to 2007 at twenty five representative hospitals by the investigators,including name,age,gender,height,weight,blood type,family history,family personnel construction,and so on.On another hand,the reviews of the medical records were done to collect all clinical information for EC patients in these hospitals.All the data of investigation were dealed using professional software.2.4 Analysis for BMI2.4.1 Computing formulaBMI was calculated with weight in kilograms divided by the square of body height in meters(kg/m~2).2.4.2 BMI gradingWe defined BMI using following criteria of BMI:GradeⅠ:underweight,BMI＜18.5;GradeⅡ:normal range weight,BMI 18.5-23.9;GradeⅢ:overweight,BMI 24.0～27.9;GradeⅣ:obesity,BMI＞28.0.2.5 The criteria for EC/GCA with and without family historyEC patients with tumor family history(EC FH+) means the families with≥2 EC patients within continuous three generations.EC patients without tumor family history(EC FH-) means the families with≤1 EC patients within continuous three generations.And at the same time,no other tumors recorded.2.6 Statistic analysisThe data were analyzed by SPSS13.0 software;Chi-squared test was performed to compare the difference between different groups.P values with less than 0.05 were considered statistically significant.Odds ratios and 95%confidence intervals(CIs) were determined from BMI and cancer case-control status.3 RESULTS3.1 Distribution of ABO blood groups in 4083 EC patients3.1.1 Characteristics of ABO blood types in Group 1,2 and 3In male EC patients,the composition percentage for blood group A in Group 1 was lower than in Group 2 and 3(25%,337/1365 vs.32%,95/617 vs.29%,70/241,P＜0.05);the percentage for blood group B in Group 1 was higher than in Group 2 and 3(38%,513/1365 vs.31%,191/617 vs.29%,70/241,P＜0.05);the percentage for blood type AB in Group 1 was a little low than in Group 2 and 3;the percentage for blood type O was a little higher than in Group 2 and 3,but the difference was not significant.In female EC patients,the composition percentage for blood type A in Group 1 was lower than in Group 2 and 3(24%,223/918 vs.30%,118/396 vs.32%,48/149,P＜0.05),the percentage for blood type B in Group 1 was higher than in Group 2 and 3 (36%,327/918 vs.33%,129/396 vs.30%,44/149,P＜0.05);the percentage for blood type AB in Group 1 was a little low than in Group 2 and 3,the percentage for blood type O in Group 1 was a little high than in Group 2 and 3,but the difference was not significant.3.1.2 Characteristics of ABO blood types in Group 4 and 5In male/female EC patients,the composition percentage for blood type A and O in Group 4 was a little high than in Group 5 respectively;the percentage for blood type B and AB in Group 4 was a little low than in Group 5 respectively,but the difference was not significant.3.1.3 Characteristics of ABO blood types in EC FH+/FH-The percentage for blood type B in EC FH+ was a little high than in FH-,the difference was not significant.The percentage for blood type O and A in EC FH+ was lower than in FH-(27%,214/803 vs.31%,454/1480;23%,187/803 vs.31%, 373/1480,P＜0.05).The percentage for blood type AB in EC FH+ was higher than in FH-(11%,89/803 vs.8%,126/1480,P＜0.05).3.2 Analysis of BMI in 1655 EC patients3.2.1 Characteristics of BMI in EC and healthy controlThe composition percentage for underweight(GradeⅠ) in EC was significantly higher than in healthy control(22%,364/1655 vs.3%,11/414,P＜0.001),the ORs was 10.330(95%CI=55.612-19.014).The percentage for normal range(GradeⅡ) in EC was a little high than in healthy control,the ORs was 1.226(95%CI=0.987-1.524). The percentage for overweight(GradeⅢ) in EC was significantly lower than in healthy control(15%,248/1655 vs.34%,140/414,P＜0.001),the ORs was 0.347 (95%CI=0.224-0.538).The percentage for obesity(GradeⅣ) in EC was significantly lower than in healthy control(3%,53/1655 vs.8%,36/414,P＜0.001),the ORs was 0.349(95%CI=0.225-0.541). 3.2.2 Characteristics of BMI in EC by genderThe composition percentage for normal range(GradeⅡ) in male EC was higher than in female EC(62%,657/1049 vs.55%,333/606,P＜0.05).The percentage for underweight,overweight and obesity in male EC was a little lower than in female respectively,but the difference was not significant.3.2.3 Characteristics of BMI in Group 1,2 and 3In EC patients,the composition percentage for underweight(GradeⅠ) in Group 1 was a little low than in Group 2 and Group 3;the percentage for normal range weight (GradeⅡ) in Group 1 was similar than in Group 2 and Group 3;the percentage for overweight(GradeⅢ) in Group 1 was a little high than in Group 2 and Group 3;the percentage for obesity(GradeⅣ) was a little high than in Group 2 and equally in Group 3,but the difference was not significant.3.2.4 Characteristics of BMI in Group 4 and 5In EC patients,the composition percentage for underweight(GradeⅠ) in Group 4 was a little low than in Group 5,the composition percentage for normal range weight (GradeⅡ) in Group 4 was a little high than in Group 5,the composition percentage for overweight(GradeⅢ) in Group 4 was a little low than in Group 5,the composition percentage for obesity(GradeⅣ) in Group 4 was a little high than in Group 5,but the difference was not significant.3.2.5 Characteristics of BMI in EC FH+/FH-The percentage for normal range BMI(GradeⅡ) was a little high than in FH-;the percentage for underweight(GradeⅠ) in EC FH+ was similar in FH-;the percentage for overweight(GradeⅢ) in EC FH+ was a little lower than in FH-;the percentage for obesity(GradeⅣ) in EC FH+ was a little high than in FH-,but the difference was not significant.3.3 Relation between ABO blood groups and BMIThe composition percentage for normal range weight(GradeⅡ) was the higher in blood group A(50%,164/331),B(44%,213/489),AB(47%,62/133) and O(41%, 144/355) than in all the other grades of BMI.The percentage for overweight(GradeⅢ) was higher in these four groups(30%,99/331;30%,148/489;30%,40/133;37%, 133/355).The percentage for obesity(GradeⅣ) in four groups of ABO was less than for overweight,with a rate of 13%(44/331),18%(88/489),15%(21/133) and 14%(51/355),respectively.The percentage for underweight(GradeⅣ) in four groups of ABO was the lowest,with a rate of 7%(24/331),8%(40/489),8%(10/133), 8%(27/355),respectively.4 CONCLUSIONS4.1 Blood type B is in higher risk for EC in Taihang Mountain area than in other regions,however,blood type A has a lower prevalence for EC,suggesting that there may be different etiological factors and mechanisms involved in esophageal carcinogenesis in these different areas.4.2 In Taihang Mountain area,blood type AB has a higher prevalence in EC FH+ than in EC FH-,blood type A and O have lower prevalence in EC FH+ than in EC FH-, suggesting that the chromosome site located the-blood type gene may harbor key genes related with EC heredity susceptibility.4.3 The males with normal range of weight(GradeⅡ,BMI 18.5-24.0)is in high risk for EC than the females,suggesting that there may be different mechanism of esophageal carcinogenesis in male and female.