Study On The Relationship Between Intake Of Folate,vitamin B12 And Vitamin B6 And Diabetes

Aim:Folate,vitamin B12 and vitamin B6 are B vitamins that play important roles in growth,development and disease prevention.At present,there are few studies on the relationship between folate,vitamin B12 and vitamin B6 intake and diabetes at home and abroad,and there are some controversies.Based on large sample data,this study used association analysis,dose-response relationship analysis,Bayesian network and other methods to analyze the relationship between intake of folic acid,vitamin B12 and vitamin B6 and diabetes.The research results can provide a scientific basis for the prevention and control of diabetes.Methods:The subjects included in this study were adults aged 20 and above in the surveys of 2007～2016.The study included questionnaires（demographic characteristics,life factors,disease history,smoking status,alcohol consumption status,and physical activity status）,laboratory tests（Fasting blood glucose,Hb A1c,2-hour blood glucose in the 75g oral glucose tolerance test）,and physical examinations（weight,blood pressure,etc.）.Dietary data were obtained through two24-hour dietary recall interviews,and diabetes and prediabetes status was determined based on self-report,drug use,and laboratory data.Student’s t-tests were used to compare the mean values between participants with and without diabetes.Chi-square tests were used to compare the percentages of categorical variables between individuals with and without diabetes.The vitamin intake is divided into two parts according to the quartile（Q1,Q2,Q3 and Q4）and whether the recommended dietary intake is met.Binary logistic regression models were used to assess the association between folate,vitamin B12,and vitamin B6 intakes and the risk of diabetes and prediabetes,respectively.Furthermore,stratified analyses were performed based on age（20-39 years,40-59 years,and≥60 years）and gender（male and female）to evaluate the relationships between folate,vitamin B12 and vitamin B6 intake and diabetes.The dose-response relationship was assessed by binary logistic regression model with the use of restricted cubic spline function with three knots located at the 5th,50th,and 95th percentiles of the exposure distribution.In the newly diagnosed population and in the prediabetes population,binary logistic regression models were used to explore the association between folate,vitamin B12,and vitamin B6 intake and new onset diabetes and prediabetes,respectively.Bayesian network model was used to study the relationship between exposure factors and the risk of diabetes from the network level,and the importance of exposure factors in each pathway was predicted.Advantage analysis was used to assess the relative importance of folate,vitamin B12 and vitamin B6 intake,and other factors on the impact of diabetes.Results:The study ultimately included 22041 adults aged 20 and over.The total weighted prevalence rate of diabetes was 18.3%,of which the male weighted prevalence rate was 19.27%and female 17.46%.1.Folate intake（1）Folate intake and diabetes:After adjusting factors such as age,gender,race,body mass index,total energy intake,vitamin B1,smoking,drinking,exercise conditions,hypertension and hypercholesterolemia,the results of binary logistic regression analysis indicated that compared with Q1,Q3 and Q4 group between folate intake and risk of diabetes association was statistically significant（P<0.05）,the OR and 95%CI were 0.62（0.48-0.79）and 0.65（0.47-0.90）,respectively,and the association was still observed in the group meeting the recommended intake compared with the group below the recommended intake（P<0.05）,OR and 95%CI was 0.69（0.56-0.86）.（2）Relationship between folate and diabetes in different age and gender:Multivariate logistic regression analysis by age and sex（Model 2）showed that folate intake was statistically significant with the risk of diabetes in people aged 20-39 years and over 60 years（P<0.05）,compared with Q1 group,the corresponding OR（95%CI）of Q4 group were 0.31（0.12-0.83）and 0.46（0.28-0.76）,respectively;In women,the association between folate intake and risk was statistically significant in both Q3 and Q4 groups compared with Q1（P<0.05）,the OR and 95%CI were 0.51（0.35-0.75）and 0.59（0.38-0.92）.（3）Relationship between folate and newly diagnosed diabetes and prediabetes:Multivariate Logistic regression analysis（Model 2）of newly diagnosed diabetes and prediabetes showed that folate intake was inversely associated with the risk of newly diagnosed diabetes in the Q3 groups compared with the Q1 group（P<0.05）,the OR and 95%CI was0.60（0.39-0.94）,while the association between folate intake and prediabetes was not statistically significant（P>0.05）;2.Vitamin B12（1）Vitamin B12 intake and diabetes:Multivariate Logistic regression analysis（Model 2）showed that compared with the Q1 group,vitamin B12 intake in the Q3group was significantly associated with the risk of diabetes（P<0.05）,OR and95%CI was 0.76（0.60-0.97）.（2）Relationship between vitamin B12 and diabetes in different age and gender:The multivariate logistic regression analysis by age and sex（Model 2）showed that the association between vitamin B12 intake and the risk of diabetes was statistically significant among people aged 60 years and older（P<0.05）,the corresponding OR（95%CI）of Q2,Q3 and Q4 groups were 0.73（0.54-0.99）,0.63（0.42-0.96）and 0.65（0.47-0.89）,respectively,compared with that of Q1group.（3）Relationship between vitamin B12 and newly diagnosed diabetes and prediabetes:Multivariate logistic regression analysis of newly diagnosed diabetes and prediabetes（Model 2）showed that compared with the Q1 group,vitamin B12intake in the Q4 group was statistically significant in association with prediabetes（P<0.05）,OR and 95%CI was 0.78（0.65-0.93）;However,the association between vitamin B12 and newly diagnosed diabetes was not statistically significant.3.Vitamin B6（1）Vitamin B6 intake and diabetes:Multivariate Logistic regression analysis（Model 2）showed that compared with the Q1 group,vitamin B6 intake in the Q4group was significantly associated with the risk of diabetes（P<0.05）,OR and95%CI was 0.61（0.42-0.89）,and the association was still observed in the group meeting the recommended intake compared with the group below the recommended intake（P<0.05）,OR and 95%CI was 0.76（0.62-0.94）.（2）Relationship between vitamin B6 and diabetes in different age and gender:Multivariate logistic regression analysis by age and sex（Model 2）showed that vitamin B6 intake was statistically significant with the risk of diabetes in people aged 20-39 years and over 60 years（P<0.05）,compared with Q1 group,the corresponding OR（95%CI）of Q4 group were 0.61（0.41-0.91）and 0.63（0.41-0.97）;In both men and women,the association between vitamin B6 intake and risk of disease was statistically significant in the Q3group compared with the Q1 group（P<0.05）,OR and 95%CI were 0.61（0.41-0.91）and 0.63（0.41-0.97）,respectively.（3）Relationship between vitamin B6 and newly diagnosed diabetes and prediabetes:Multivariate Logistic regression analysis（Model 2）showed a statistically significant association between vitamin B6 intake and the risk of newly diagnosed diabetes in the Q3 groups,as compared to the Q1 groups（P<0.05）,and the OR and 95%CI were0.65（0.43-0.98）.Compared with the Q1 group,vitamin B6 intake in the Q3 and Q4group was statistically significant in association with prediabetes（P<0.05）,OR and95%CI were 0.80（0.66-0.98）and 0.72（0.58-0.89）.4.Dose response relationship between vitamin intake and risk of diabetesThere was a nonlinear negative association between dietary folate intake and diabetes(P_nonlinearity<0.01),and no significant association was found with an intake beyond 1098 mcg/day（OR:0.70;95%CI:0.49–1.00）.Meanwhile,a linear negative association between vitamin B12 intake and the risk of diabetes was also found(P_nonlinearity=0.051).With an increase of vitamin B12 intake,there was no significant association in diabetes risk beyond 22 mcg/day（OR:0.74;95%CI:0.55–1.00）.In addition,there also existed a nonlinear negative association between vitamin B6intake and the risk of diabetes(P_nonlinearity<0.01),and the prevalence of diabetes decreased with increasing vitamin B6 intake and reached a plateau when vitamin B6intake above 6 mg/day（OR:0.70;95%CIs:0.53-0.92）.5.Bayesian network analysisThrough construct a bayesian network analysis contains nine variables（including movement status,smoking status,alcohol consumption condition,high blood pressure,high cholesterol,folate,vitamin B12,vitamin B6,and diabetes）of bayesian network,including the impact of diabetes five direct parent node,there are five main variables between interdependent pathway.Alcohol consumption and hypercholesterolemia are the key factors affecting diabetes mellitus.The results of prediction of diabetes influencing factors by a Bayesian network model show that hypertension is the most influential factor,while vitamin B6 is the least influential factor.6.The dominance analysis,according to the results of the standardization of diabetes factors affecting the weight advantage from big to small,followed by age（28.50%）,hypertension（21.04%）,BMI（18.11%）,hypercholesterolemia（11.80%）,physical activity status（6.23%）,race（4.27%）,energy intake（4.03%）,and smoking status（2.06%）in descending order.,gender（1.99%）,drinking status（1.73%）,folate intake（0.15%）,vitamin B6 intake（0.06%）,vitamin B12 intake（0.02%）,and vitamin B1 intake（0.01%）.Conclusions:Increased intakes of folate,vitamin B12 and vitamin B6 may be associated with a reduced risk of diabetes.Appropriate increases in intake of folate,vitamin B12 and vitamin B6 are recommended to reduce and prevent the occurrence of diabetes.