Derivation Of Genetic Models Based On The Association Between Rs2241766 And Rs266729 Locus And Type 2 Diabetes Mellitus

ObjectiveThis study was conducted to estimate the genetic effects of the rs2241766,rs266729 on type 2 diabetes mellitus?T2DM?and the underlying genetic models of inheritance,which could solve the problem caused by unknown genetic model and reasonably evaluate the association between genetic markers and diseases.MethodPotential eligible studies about the association between polymorphisms of adiponectin gene and T2DM,were identified by systematically searching databases in home and abroad for original articles of prospective cohort,case-control or whole genome association studies with necessary data.The following information was extracted from all included studies:author,year of publication,continent,country,nationality,population,standards of DM diagnosis,average age and standard deviation,and frequencies of genotypes in both groups,fasting plasma glucose?FPG?,etc.Based on the information extracted from studies,the generalized odds ratio?GOR?value was estimated to assess the association between the G alleles of the two loci and T2DM under complete genotype distribution without a genetic model in ORGGASMA software,then the genetic model was estimated.The parameter?was calculated to estimate the genetic model according to the study of John H.Gillespie.If?was 0,1,or between 0 and 1,the corresponding genetic model of the G allele was recessive,dominant or incompletely dominant,respectively;If?>1 or<0,this indicated over-dominance or under-dominance.When?was 0.5,it was called additive,a type of special incomplete dominance model.Genetic model estimation of qualitative data:the quotient of natural logarithm odds ratio?LogOR₁?of MW to WW?M mutant gene loci,W wild gene loci?divided by natural logarithm odds ratio?LogOR2?of MM to WW was defined as parameter?,where LogOR₁ was the dependent variable?outcome variable?and LogOR₂ was the independent variable?non constant?.The genetic effect was estimated by binary logistic regression analysis based on the genetic model,where the genotypes being amalgamated or assigned were defined as independent variables,the groups?T2DM cases and controls?were defined as binary dependent variables.Genetic model estimation of quantitative data:Mean difference?MD?was calculated between T2DM group and control group.The quotient of MD₁ of MW minus WW divided by MD₂ of MM minus WW was defined as parameter?,where MD₁ was the dependent variable and MD₂ was the independent variable?non constant?.The genetic effect was estimated by linear regression analysis based on the genetic model,and the genotypes were defined as independent variables,FPG was defined as dependent variable.The statistical analysis for genetic model and genetic effect estimation was completed in Stata 12?Stata Corporation,USA?,the significance level was set at0.05.Results1.Qualitative outcome?T2DM or not?associated with rs2241766 were analyzed.The G allele was a risk factor for T2DM because GOR was 1.19 and 95%confidence interval?CI?was?1.06,1.35?.The subjects were divided into two groups according to the continent:Asian population,European and American population.Asian:the results of genetic model estimation showed that confidence interval including 0.5,?was 0.5,and the genetic model for the effect of the G allele on T2DM was additive??=0.43,95%CI?0.30,0.55?,P<0.001?.European and American:G allele has no relationship with T2DM and the genetic model could not be determined??=-0.09,95%CI?-0.44,0.27?,P=0.63?.The result of genetic effect estimation based on additive genetic model in Asian showed that:odds ratio?OR?was 1.12,95%CI was?1.07,1.18??P<0.001?,indicating that the risk of T2DM for the later increases by 12%than the former as the sequence of TT?GT?GG genotype.Qualitative outcome associated with rs266729 were analyzed.The G allele was a risk factor for T2DM because GOR was 1.19 and 95%CI was?1.05,1.23?.The subjects were divided into three groups according to the continent:Asian population,European and American population,African population.African sample was too few to be analyzed.Asian:the results of genetic model estimation showed that confidence interval including 0.5,?was 0.5,and the genetic model for the effect of the G allele on T2DM was additive??=0.36,95%CI?0.13,0.60?,P<0.001?.European and American:the genetic model for the effect of the G allele on T2DM was incompletely dominant??=0.19,95%CI?0.03,0.35?,P<0.001?.The result of genetic effect estimation based on additive genetic model in Asian showed that:OR was 1.11,95%CI?1.06,1.17??P<0.001?,indicating that the risk of T2DM for the later increases by 11%than the former as the sequence of CC?GC?GG genotype.There were no statistical significance in European based on incompletely dominant genetic model?P>0.05?.2.Quantitative outcome associated with rs2241766 were analyzed,all subjects were Asian.The average FPG of TT,GT and GG genotypes in controls were 5.58,5.79 and6.64?mmol/l?,respectively;the average in cases were 8,67,8.99 and 9.23?mmol/l?,respectively.The genetic model for the effect of the G allele on controls didn't need to be estimated.The result of genetic model estimation in cases showed that confidence interval including 0.5,?was 0.5,and the genetic model for the effect of the G allele on T2DM was additive??=0.39,95%CI?0.14,0.64?,P=0.002?;and there was no statistical significance in linear regression analysis based on additive genetic model??=0.45,t=1.18,95%CI?-0.36,1.25?,P=0.26?.There were no rs266729 quantitative data collected.Conclusion1.The genetic model for the effect of the G allele of rs2241766 on T2DM in Asian qualitative data was additive;the personals who have G allele will have13.0%higher risk for T2DM than personals who have not G alleles;the genetic model in European could not been determined.2.The genetic model for the effect of the G allele of rs266729 on T2DM in Asian qualitative data was additive,the personals who have G alleles will have11.0%higher risk for T2DM than personals who have not G alleles;the genetic model in European was incomplete additive.3.The parameter?calculated for genetic model estimation and linear regression for genetic effect estimation?no constant?were unified for qualitative and quantitative data analyses.