Risk Factors And Mechanism Of Cancer In Type 2 Diabetes Patients In Changchun

Background:Epidemiological studies have shown that tumorigenesis is more common in diabetic patients than in non-diabetic patients.To investigate the risk of cancer in patients with type 2 diabetes mellitus(T2DM),analyze risk factors and further elucidate the mechanisms of these relationships to identify potential targets for screening and intervention.Purpose:This study intends to use a nested case-control study design in a retrospective cohort to screen the risk factors for malignant tumors in patients with type 2 diabetes,construct a prediction model using classical machine learning algorithms,and evaluate the relationship between diabetes and its course,related clinical indicators,and hypoglycemic therapeutic drugs and cancer risk,to provide clinical recommendations for the prevention and management of type 2 diabetes and cancer.Secondly,using the baseline epidemiology and follow-up diabetes information in the Changchun health cohort established from 2011 to 2021,the basic information of the study subjects was collected and whether T2DM was an independent risk factor for tumor development was analyzed.Subsequently,the GEO database was used to explore the possible mechanism of tumor co-occurrence in T2DM patients from the perspective of multiple genes,and the horizontal landscape of immune cells in T2DM and malignant tumors was initially constructed to explore the association between immune cells and T2DM co-occurrence of cancer.Finally,a two-sample Mendelian randomization method was used to explore the causal association between T2DM and blood glucose-related indicators and malignant tumors.Methods:1.The subjects of the case-control study were 46066 patients with type 2 diabetes who were admitted to the inpatient department of endocrinology and metabolism of 4 first-class hospitals in Changchun,Jilin Province from 2013 to 2022.They were divided into the T2DM group(n1=30272)and the T2DM group with tumor(n2=15794).General demographic data,diabetes course,personal history,disease history,blood glucose,and lipid-related indicators,circulating trace element levels,and insulin resistance were compared between groups.Single-factor analysis was used to screen the risk factors of tumors in T2DM patients.2.The relationship between the use of hypoglycemic drugs and tumor risk in T2DM patients was analyzed:the number of patients taking each drug,the number of patients using each drug alone,and the number of patients taking combined drugs were counted respectively.After adjusting for confounding factors such as age,gender,and medical history,the relationship with concomitant malignant tumors was analyzed using multivariate logistic regression.Insulin use was divided into three dose groups,low,medium,and high,and the difference in the risk of concomitant tumor was analyzed.3.The subjects of the cohort study were from the healthy cohort of Changchun from 2011 to 2021,a total of 9529 adults,including 1501 adults with T2DM and 8028 adults were non-diabetic population.The study was used to analyze the baseline data of type 2 diabetic patients,and Cox proportional risk regression model was used to analyze whether T2DM was an independent risk factor for cancer,and the impact of lifestyle,emotional state,and disease history on cancer risk of T2DM patients was explored.4.Construction of cancer risk prediction models for type 2 diabetes patients:LASSO algorithm was used to screen characteristic variables,logistic regression,random forest,and support vector machine algorithms were used to construct cancer risk prediction models for type 2 diabetes patients,and Delong test was used to compare the prediction effects of the models.5.Mechanism exploration of T2DM and tumor risk:Sequencing data of cancer tissues and adjacent tissues,as well as gene expression data of T2DM samples and normal samples,were selected for ten types of cancer with a high incidence in T2DM patients in the GEO database.Combining datasets of the same type of tumor,Using|log2foldchange|>1 and p.AJJ<0.05 as the screening criteria,CDEGs(Common differential expression genes)of different cancers and T2DM differences were screened.Functional enrichment analysis and PPI network construction were conducted for CDEMs in gastric cancer and colorectal cancer.The top 10 most important hub genes were selected by the maximum Clique Centrality(MCC)algorithm using the CytoHubba plug-in.Subsequently,the Cibersort software package was used to analyze the population abundance of infiltrating immune cells and stromal cell populations and Spearman correlation analysis was used to calculate the correlation between common DEM expression and the abundance of 22 immune infiltrating cells,and correlation heat maps were drawn.6.The two-sample Mendelian randomization method was used to explore the causal association between T2DM and blood glucose-related indicators and gastric cancer and colorectal cancer:FPG,2h-PG,fasting insulin(FIRI),HbAlc,and T2DM were used as exposure factors,with genetic prediction data for glucose traits from the 2021 GWAS study conducted by the Glucose and Insulin-Related Traits Consortium(MAGIC),adjusting for age,study location,and maj or components as covariates.SNPs outcome data for gastric cancer were obtained from a large GWAS study conducted by the Biobank of Japan(BBJ).SNPs sites that reached the genome-wide significance threshold(p<5×10-8)were used as genetic tool variables,and the threshold was widened to p<5×10-6 when only 2 SNPS or less were obtained under this threshold.Horizontal pleiotropy was evaluated by the MR PRES SO method and MR-Egger regression intercept.Results1.A total of 46066 T2DM patients were included in this study,including 30272 T2DM patients alone and 15794 T2DM patients with tumors(T2DM+CA).Patients in the T2DM+CA group had a higher proportion of females,were older,had a higher BMI,and had a longer duration of diabetes(p<0.001).2.Among the malignant tumors associated with T2DM,digestive system tumors accounted for 4242 cases(32.14%),followed by respiratory system 3043 cases(25.78%),genitourinary system 3131 cases(23.72%),endocrine and immune system 1080 cases(8.18%),blood and hematopoietic system 589 cases(3.46%),etc.In addition to the gender specificity of male and female reproductive system malignancies,the incidence of digestive system malignancies in men(40.13%),respiratory system malignancies in men(30.28%),lymphatic system malignancies in men(3.93%),blood and hematopoietic system malignancies in men(5.57%)was higher than that in women(p<0.05).3.The risk of malignant tumors in patients with diabetes in different age groups:the age of patients with tumors of different systems was significantly higher than that in T2DM alone group(all p<0.001).As people with type 2 diabetes age,the risk of developing malignant tumors increases gradually and significantly.Compared with patients younger than 20 years of age,patients with diabetes aged 20-30,30-40,40-50,50-60,60-70,70-80,and>80 years of age had a 1.43,2.03,2.86,4.24,5.90,6.00,and 7.94 times greater risk of developing malignant tumors,respectively.4.The influence of BMI on malignant tumors in type 2 diabetic patients:Compared with patients with BMI ≤24kg/m2,patients with type 2 diabetes with a BMI of 24-28 kg/m2,28-32 kg/m2,and>32 kg/m2 were 1.53,1.38,and 1.21 times more likely to develop malignant tumors than patients with BMI≤24kg/m2,respectively.5.The impact of diabetes course on the development of malignant tumors in patients with diabetes:The risk of tumor development was highest in patients 5 to 10 years after diabetes diagnosis,which was 2.51 times that of patients with diabetes course<5 years.The risk of concomitant malignancy decreased with increasing duration of diabetes but was still significantly higher than that of T2DM patients with a duration less than 5 years(p<0.001).6.Daily insulin dose was an independent risk factor for concurrent cancer in T2DM patients,and the cancer risk was 1.481 and 1.808 times higher in the medium-dose and high-dose groups than in the low-dose group,respectively(p=0.003).7.The correlation analysis results between tumor markers and blood glucose indicators showed that CA199,CA125,CA153,CA50,CEA and cPSA were significantly positively correlated with FPG.CA199,CA153,CEA,cPSA,SF were significantly positively correlated with 2h-PG.There was a significant positive correlation between APT and INS.CEA and cPSA were significantly negatively correlated with CP,while HE-4 was significantly positively correlated with CP.CA199,CA242,CEA,CA125,AFP,cPSA and HbAlc show significant positive correlation.CA199 was also positively correlated with HOMAIR(all p<0.05).8.Baseline data from the cohort study design showed that people with T2DM were older,heavier,had wider waist and hip circumference,and had higher BMI and waistto-hip ratio compared with controls.The prevalence of T2DM was the highest in the age group of 60-69 years old,the prevalence of diabetes increased with age before 70 years old,and the prevalence of T2DM after 70 years old had a decreasing trend.Weight tend to decrease with age,while waist circumference and waist-to-hip ratio tend to increase.The age group of 60～69 years had the highest BMI and the highest prevalence.Family smoking,exposure to chemicals and radiation are risk factors for T2DM.Abstaining from alcohol and occasional tea consumption are protective factors for T2DM.The number of nights eaten per week>3,the number of breakfast days per week<3,and the number of lunch days per week<3 were risk factors for T2DM.Being mildly active is a protective factor for T2DM and also for tumor risk in T2DM patients.Myocardial infarction,stroke,coronary heart disease,hypertension,lower extremity arterial disease,hyperlipidemia,cholecystitis,gallstones,chronic pancreatitis,and kidney stones are risk factors for T2DM.9.Cox proportional regression analysis showed that T2DM was an independent risk factor for cancer risk after adjusting for age,sex,BMI,and medical history.Patients who follow the doctor’s advice to control their diet,regularly take hypoglycemic drugs,regularly use insulin,and supplement nutrients are protective factors for cancer in patients with T2DM.Almost daily thoughts of suicide or self-harm are risk factors for cancer in T2DM patients,and the risk of cancer in T2DM patients with poor family relationships is higher than that in patients with good family relationships.10.A restricted cubic spline analysis adjusting for age,sex,BMI,and disease history showed a "J-shaped" association between FPG and 2h-PG and cancer risk.The lowest cancer risk was 5.1 mmol/L FPG and 8.1 mmol/L 2h-PG.In addition,cancer risk decreased as FPG increased in the range of(2.4-5.1)mmol/L,but the risk of concomitant cancer began to rise significantly as FPG continued to increase beyond 5.1 mmol/L.As 2h-PG increases in the range of(4.9-8.1)mmol/L,the risk of cancer decreases,but as 2h-PG continues to increase beyond 8.1 mmol/L,the risk of concomitant cancer begins to rise significantly.11.By diabetes course,gender,BMI,insulin dosage,The malignant tumor risk prediction model of T2DM patients with smoking,drinking,FPG,FIRI,HbAlc,CPHD(cancer-protective hypoglycemic drugs)and CRHD was constructed with a total of 11 variables as predictors.logistic regression model,random forest model,and support vector machine model all have good prediction effects.The random forest model has the best prediction effect,and the area under the ROC curve reaches 0.9455.12.The differential genes between tumor tissue and paracancer in gastric cancer share the most common differential genes with T2DM DEGs,with 106.This was followed by lung cancer with 105,breast cancer 101,colorectal cancer 74,cervical cancer 45,prostate cancer 43,liver cancer 42,bladder cancer 26,endometrial cancer 24 and pancreatic cancer 21.CDEGs of gastric cancer and T2DM are significantly enriched in transcription disorders in cancer,diabetic gastrointestinal disease,and IGF1,NF-kB,JNK/SAPK,AMPK and other pathways.Among them,MYC and COL3A1 are the key genes in the common differential genes of gastric cancer and T2DM.CDEGs of colorectal cancer and T2DM are significantly enriched in the mineral absorption pathway,among which MT1 gene family,PLA2G2A,etc.are key genes in the common differential genes of gastric cancer and T2DM.13.The results of the two-sample Mendelian randomization study showed that IVW,MR-Egger and WME analysis methods all suggested a causal relationship between T2DM and gastric cancer and colorectal cancer.It did not suggest a causal association between FPG,FIRI,HbA1c,2h-PG and the two cancers.Sensitivity analysis showed that no SNP would have a bias effect on the results,and the results were relatively robust.Conclusions:1.The risk of malignancy is higher in patients with type 2 diabetes who are older,have a higher BMI,have a longer duration of diabetes,have a smoking and drinking habit,and have a history of coronary heart disease or stroke.Especially for T2DM patients older than 40 years,with a BMI of 24-28 kg/m2 and a diabetes course of 5-10 years,it is necessary to conduct tumor screening.2.For the use of antidiabetic drugs in type 2 diabetic patients with advanced age,high BMI,long duration of diabetes,history of smoking and drinking,coronary heart disease,and stroke in terms of the risk of cancer,it is recommended to choose antidiabetic drugs with the effect of weight control,rather than insulin secretagogues and high doses of insulin.3.T2DM is an independent risk factor for cancer risk,with a cancer incidence of 4.3%in the diabetes group,significantly higher than 0.8%in the control group.T2DM patients should follow the doctor’s advice to control diet,regularly take hypoglycemic drugs,do light activities,reduce the number of midnight snacks,eat breakfast and lunch regularly,and maintain physical and mental health.Especially for T2DM patients with myocardial infarction,stroke,coronary heart disease,hypertension,lower extremity arterial disease,hyperlipidemia,cholecystitis,gallstones,chronic pancreatitis,kidney stones,they are high-risk groups of cancer,and should pay attention to the adjustment of the above lifestyle habits.4.FPG and 2h-PG showed a "J"-shaped association with cancer risk.The lowest cancer risk was 5.1 mmol/L FPG and 8.1 mmol/L 2h-PG.Clinically,the FPG and 2hPG indexes of T2DM patients at high risk of tumor should be closely observed.In view of their nonlinear correlation with tumor,either too high or too low will increase the risk of tumor.5.The Random forest model for predicting malignant tumor risk in T2DM patients,which was constructed with 11 variables as predicators,including diabetes course,gender,BMI,insulin dosage,smoking,drinking,FPG,FIRI,HbAlc,CPHD,and CRHD,had a good predictive effect.6.Multiple tumor markers were significantly correlated with blood glucose or insulin resistance related indicators such as FPG,2h-PG,INS,CP,HOMAIR,etc.It is suggested that T2DM may be associated with the pathogenesis of malignant tumor.It may be a risk factor for tumor in patients with type 2 diabetes,and close observation of the above indicators can provide early signals of tumor in patients with type 2 diabetes.7.The common differential genes of gastric cancer and T2DM may participate in the pathogenesis of gastric cancer associated with T2DM by regulating signaling pathways such as transcription dysregulation in cancer and diabetic gastrointestinal disease.The common differential genes of colorectal cancer and T2DM may be involved in the pathogenesis of colorectal cancer in T2DM patients by regulating mineral absorption and other pathways.Two-sample Mendelian randomization suggests a causal relationship between T2DM and both gastric and colorectal cancer.