Diabetes Mellitus Is Associated With The Increased Risk Of Colorectal Cancer

1.Background and PurposePrevious epidemiological studies indicated DM was associated with the increased risk for CRC and other cancers although no entirely consistent results were made. Previous studies have been limited by incomplete characterization of DM status, variably defined control population,etc.Organ- specific cancer risks remain incompletely defined and results are inconsistent.In the present study,we examined the associations between the prevalence of diabetes and incidence of CRC with an international ecological study and populationbased risk analysis for CRC by overall and by anatomic sub-site among type 2 diabetes patients in China.2.Methods2.1 Ecological studyThe prevalence data of diabetes in 2000 all over the world was obtained from the diabetes database of International Diabetics Federation.Diabetic prevalence data for adults(≥20 years of age) were originated from a defined,population-based sample and diagnosis of diabetes based on optimal WHO criteria.The specific incidence rates of CRC in 2002 all over the world were obtained from GLOBOCAN 2002 database.The GLOBOCAN 2002 database has been built up using the huge amount of data available in the Descriptive Epidemiology Group of International Agency for Research on Cancer(IARC).Incidence data were available from population- based cancer registries.They covered entire national populations,or samples of such populations from selected representative regions.The incidence rates were standardized based on the world standard population.The diseases considered in the present study were colon and rectum cancer(C18-C21,International Classification of Disease(ICD,10th revision)).The original prevalence rate of diabetes and incidence rates of CRC were square root- transformed.Scatter plots were used to examine the rough relations between the prevalence rates of diabetes and the incidence rates of CRC.Then Spearman's correlation coefficients were calculated if the trends of associations were indicated. Linear regression analysis was carried out with transformed prevalence of diabetes as independent variable and the incidence rate of CRC as dependent variable.All the calculations and graphs were performed with SPSS software system(SPSS Inc., Chicago,IL,USA).A P value less than 0.05 was considered as statistically significant.2.2 Population- based risk studyThe field of population- based study was at Nan-Hu district of Jia-Xing city, Zhejiang Province,China,which is situated at the south-east China and one of highest incident region of CRC in China.Subjects in this study were the recorded patients with type 2 diabetes(ICD 10 code E11) in Nan-Hu district during 2002.1.1-2008.6.30.Person-years of follow-up for each subject was calculated from the date that type 2 diabetes was first diagnosed to the date of incident CRC diagnosis,death,or June 30,2008,whichever occurred first.CRC in all the subjects were identified through record-linkage DSRD with CSRD.Colon cancer cases comprised codes 18.0 to 18.9 and rectal cancer cases comprised codes 19 and 20 by the International Classification of Disease(ICD,10th revision).Overall,gender- and sub-site specific incidence rate of CRC were calculated by dividing the number of incident cases by observed person-years.95%confidence intervals(95%CI) for incidence rates were calculated based on the Poisson distribution. Standardized incident ratio(SIR) and its 95%CI were used to estimate the risk for CRC incidence in the subjects with type 2 diabetes.SIR was calculated as the ratio of the number of observed CRC cases to the number of expected CRC cases.The number of expected CRC cases was calculated based on the age-,gender- specific CRC incidence rates of Nan-Hu general residents as the standard incidence rate during the same period.The databases were set up using Microsoft Excel software and all the calculations were carried out using SPSS software system(SPSS Inc.,Chicago,IL,USA).3.Results3.1 Correlation between the prevalence of diabetes and the incidence of CRC170 countries had both prevalence of diabetes and incidence of CRC data available for analysis.Positive correlations were found between the prevalence of diabetes and the incidence of CRC.The spearman's coefficients of correlation(rs) in male and female were 0.534 and 0.597(Both of P value <0.001),respectively.Using analysis with linear regression model,significant linear associations were found in male and female subjects. The prevalence of diabetes was the significant positive predicator of CRC incidence.3.2 The risk for CRC among type 2 diabetic patientsDuring 2002.1.1-2008.6.30,7938 type 2 diabetic patients were reported through Diabetics Surveillance and Registry Database system in Nan-Hu district,Jia-Xing City. Among 7938 type 2 diabetic patients,3792(47.8%) were men and 4146(52.3%) were women.The mean age was 61.1 years and 9.98%diabetic patients had the family histories of diabetes.The total number of observed time was 20 331.5 person years. During this period,64 cases of CRC in 7938 type 2 diabetic patients were identified through the record-linkage DSRD with CSRD.The overall crude incidence rate of CRC in diabetic patients was 314.8(95%CI:237.6-391.9) per 10~5 person years.Among 64 CRC patients,44 cases are colon cancers and 20 cases are rectal cancers.The colon cancer incidence rate among type 2 diabetes was 216.4(95%CI:152.5-280.4) per 10~5 person years and 98.4(95%CI:55.1-141.6) per 10~5 person years for rectal cancer.The proportion of colon cancer in these 64 CRC patients was 68.8%,which was higher than that in general CRC patients(59.1%).However,the difference of the proportion of colon cancer did not arrive statistical significance(P>0.05).With the age-specific incidence rate of CRC of the general population in the surveillance region as the standard incidence rate,SIRs for CRC among type 2 diabetic patients by gender and sub-type of disease were calculated.SIR for CRC among diabetic subjects was 158.8(119.9-197.7).Type 2 diabetic patients increased about 58.8%of risk compared with the general population.Stratified by gender,SIR for CRC among male patients was 182.1(95%CI:123.4-240.8) and was more evident than that among female patients 136.4(85.0-187.9).No significant association was found in female type 2 diabetic patients.Stratified by anatomic sub-type of CRC,significant increased risk for colon cancer was found with SIR of 179.0(126.1-231.9).Male type 2 diabetic patients showed a significant risk for colon cancer with the SIR of 189.9 (113.9-265.8).Female type 2 diabetic patients showed a borderline significant risk for colon cancer with the SIR of 168.7(94.8-242.6).However,no significant associations were found between diabetes and risk for rectal cancer among male(172.3(78.6-266.0)) and female(90.6(23.5-157.8)) patients(all the P value >0.05).4..ConclusionsDiabetes was associated with the increased risk for CRC and this association was more evident for colon cancer and among male diabetic patients.