| Increasing prevalence of type 2 diabetes worldwide has become a serious public health problem.Type 2 diabetes would cause diabetic nephropathy,cardiovascular and cerebrovascular diseases,diabetic eye diseases,and diabetic foot,even death,having tremendous deleterious influences on people's health.Type 2 diabetes and its complications related to huge health costs placed enormous pressure on individuals and the global health economy.Endocrine disrupting chemicals(EDCs)is a kind of environmental pollutants with endocrine disrupting effects,including Bisphenol A(BPA),which belongs to the typical alkyl phenolic EDCs,Perfluorooctanoic Acid(PFOA)and Perfluorooctane Sulfonates(PFOS),which belongs to typical perfluorinated alkyl EDCs.BPA,PFOA,and PFOS are widely used in modern industrial and commercial applications.The main exposure pathways of BPA,PFOA,and PFOS were diet,breathing,and skin contact.BPA,PFOA,and PFOS exposure may cause adverse health effects.In recent years,associations between BPA,PFOA,and PFOS and the risk of diabetes has attracted extensive attention.However,most of the studies limited to small sample size cross-sectional studies with inconsistent conclusions.In addition,humans are commonly exposed simultaneously to multiple EDCs,and multiple EDCs may interfere with each other affecting their effects on type 2 diabetes.However,most of the previous studies focused on the association of single EDCs with type 2 diabetes risk.Therefore,in the present study,a prospective nested case-control study was conducted to investigate the association between serum BPA,PFOA,and PFOS and type 2diabetes risk,and further to explore the associations between serum BPA,PFOA,and PFOS and type 2 diabetes risk in the case of simultaneous exposure of these three EDCs.Genetic factors and their interaction with environmental factors played an important role in the development of type 2 diabetes.Differences in genetic background might contribute to the inconsistent findings of the associations between EDCs and type2 diabetes.To the best of our knowledge,only one study examined the interaction of genetic risk score(GRS),constructed with 34 diabetes-related single nucleotide polymorphisms(SNPs),and urinary BPA on the type 2 diabetes.risk This study neither reported a significant association of urinary BPA with type 2 diabetes,nor modification effect of GRS on the association between them.However,the relatively small sample size in the type 2 diabetes cases and limited diabetes-related SNPs included in this study,might result in insufficient statistical power.Meanwhile,no study investigated the interaction of diabetes GRS and serum PFOA and PFPS on type 2 diabetes risk.Therefore,more studies with larger sample size and more SNPs are needed to explore the effects of genetic factors on the association between EDCs and the type 2 diabetes risk and to provide the scientific evidence for accurate prevention in the diabetic high-risk population.Moreover,the mechanism of type 2 diabetes development induced by BPA,PFOA,and PFOS exposure is not fully understood.BPA is an estrogen receptor agonist that acts as an endocrine hormone disruptor and is involved in a variety of mechanisms of the development of diabetes,including glucose homeostasis,obesity,insulin resistance,abnormal ? cell function,inflammation,and oxidative stress.PFOA and PFOS are similar in structure to fatty acids and play a key role in the regulation of adipocyte differentiation,lipid and glucose metabolism,and inflammation by activating peroxisome proliferation activation receptors.Therefore,based on the prospective dynamic Dongfeng-Tongji(DFTJ)cohort,the present study adopted a nested case-control design,including 995 incident type 2diabetes cases during the first follow-up period from 2008 to 2013.Matched 995controls(1:1 matched on age and gender)were randomly selected from participants free of coronary heart disease,stroke,or tumor at baseline and follow-up visits.Demographic data,lifestyle,disease history,and anthropometric data at baseline and the first follow-up were collected by questionnaire and physical examination.Serum BPA,PFOA,and PFOS concentrations were measured in the laboratory.Diabetes GRS was constructed by 88 diabetes-related SNPs obtained from GWAS and meta-analyses of large GWAS.The present study firstly explored the distribution of BPA,PFOA,and PFOS and their associations with diabetes risk factors,and then investigated the associations of serum BPA,PFOA,and PFOS with type 2 diabetes risk,respectively.Further,we explored the effects of serum BPA,PFOA,and PFOS on the 5-years changes of FPG in participants without diabetes and diabetes cases.We also examined the interaction and combined effects of diabetes GRS and serum BPA,PFOA,and PFOS on the risk of type 2 diabetes.Serum BPA,PFOA,and PFOS were included simultaneously in models to investigate the association between the three EDCs and the risk of type 2 diabetes.In detail,the present study consists of the following three parts:Part I: Associations between serum BPA,diabetes genetic risk score,and type 2 diabetes riskObjective: To investigate the association of serum BPA with the type 2 diabetes risk,and the interaction and combined effects of diabetes GRS and serum BPA on the risk of type 2 diabetes.Methods: In the present study,995 newly diagnosed diabetes cases in the DFTJ cohort during the first follow-up from 2008 to 2013 were included.Then,995 controls(1:1 matched on age and gender)were randomly selected from participants free of coronary heart disease,stroke,or tumor at baseline and follow-up visits.Type 2 diabetes was diagnosed based on the following criteria: fasting blood glucose ? 7.0mmol/L,or glycosylated hemoglobin ? 6.5%,or use of hypoglycemic agents,or self-reported physicians-diagnosed type 2 diabetes.Serum BPA concentration at baseline was measured by high performance liquid chromatography coupled with mass spectrometry(HPLC-MS).Diabetes GRS and weighted GRS(w-GRS)was constructed by 88 diabetes-related SNPs selected from large GWAS and the meta-analysis of GWASs.Conditional Logistic regression model was used to estimate the association between serum BPA and the risk of type 2 diabetes.The interaction and combined effects of GRS and BPA on the risk of type 2 diabetes were analyzed by unconditional Logistic regression model.Generalized linear model was used to examine the association of serum BPA with 5-years FPG changes,and the potential modification effects of diabetes GRS on the association between them.Results: The median of serum BPA concentration was 3.08ng/m L and 2.81ng/m L in incident cases and controls,respectively.In the analysis of the associations between serum BPA and diabetes risk factors,serum BPA concentration was significantly higher in males than in females(3.15ng/m L vs.2.83ng/m L,P = 0.02).Serum BPA concentration was negatively correlated to waist circumference(? =-0.0033,P = 0.01).Participants without hypertension had higher serum BPA concentration than those with hypertension(3.12ng/m L vs.2.68ng/m L,P = 0.001).Serum BPA concentration was positively associated with total cholesterol(TC)(? = 0.0358,P = 0.01),low density lipoprotein(LDL)(? = 0.0507,P = 0.02),and FPG(? = 0.0635,P = 0.0008).Serum BPA concentration was significantly higher in subjects with FPG higher than 5.6mmol/L than in those with normal FPG(3.34ng/m L vs.2.74ng/m L,P = 0.0001).Serum BPA levels were not correlated with age,BMI,education level,diet,hyperlipidemia,and triglyceride(TG)levels(all P > 0.05).Participants were classified into tertiles based on the serum BPA concentration in the control group.Compared with tertile 1,the odds ratio(OR)of tertile 3 was 1.07 [95% confidence interval(CI),0.85-1.35,Ptrend = 0.86].After adjusted for traditional type 2 diabetes risk factors,the corresponding OR value was 1.10(95%CI,0.72-1.66 Ptrend = 0.78).In stratified analysis,no interactions of serum BPA levels and diabetes risk factors on type 2 diabetes risk were detected.(all Pinteraction > 0.05).Both GRS and w-GRS were significantly associated with the risk of type 2 diabetes(? = 0.040 and 0.366,P < 0.01),but there was no interaction with serum BPA levels(all Pinteraction > 0.05).In the analysis of the combined effects of GRS or w-GRS and serum BPA levels on type 2 diabetes risk,participants were classified into tertiles based on the distribution of GRS and w-GRS,the ORs of the group with both highest GRS or w-GRS and serum BPA concentration were 2.37(95%CI,1.37-4.08,P = 0.002)and 2.37(95%CI,1.36-4.14,P = 0.002),respectively,compared with those with both lowest GRS or w-GRS and serum BPA concentration group.In both incident cases and control groups,serum BPA levels were not significantly associated with 5-years FPG changes(P > 0.05),and GRS or w-GRS did not modify the association between them.Conclusions: Males,subjects with larger waist circumference,no history of hypertension,higher TC,LDL,and FPG had higher serum BPA concentration.The present study did not find significant association between serum BPA and the risk of type 2 diabetes and 5-years FPG changes.There was no modification effect of GRS on the association of serum BPA with the risk of type 2 diabetes and 5-years FPG changes.Combined effects of GRS on the association of serum BPA with type 2 diabetes risk were observed.Although the present study is the largest prospective study,more prospective studies with larger sample size and more SNPs are warranted to confirm our findings in the future.Part II: Associations between serum PFOA,PFOS,diabetes genetic risk score,and type 2 diabetes riskObjective: To investigate the association of serum PFOA and PFOS with the type 2 diabetes risk,and the interaction and combined effects of diabetes GRS and serum PFOA and PFOS on the risk of type 2 diabetes.Methods: Participants' characteristics,the definition of type 2 diabetes,and GRS construction were the same as the first part.Baseline serum PFOA and PFOS concentrations were measured by HPLC-MS.In the analysis of the correlation between diabetes risk factors and serum PFOA and PFOS,Spearman correlation was used to examine the correlation between continuous variables and serum PFOA and PFOS concentrations(lg conversion),and generalized linear model adjusted for age and gender was used to detect the associations between categorical variables and serum PFOA and PFOS.The associations between serum PFOA and PFOS and the risk of type 2 diabetes was determined by conditional Logistic regression.PROCESS plug-in in SPSS was used to estimate the mediating effects of blood lipid,FPG,and BMI on the associations between serum PFOA and the risk of type 2 diabetes.The interaction and combined effects of GRS and serum PFOA and PFOS on the risk of type 2 diabetes were assessed by unconditional Logistic regression.Associations and interaction effects of GRS and serum PFOA and PFOS on type 2 diabetes risk and 5-years FPG changes were estimated by generalized linear model in incident cases and controls.Results: The median serum PFOA concentration was 1.31 and 1.14ng/m L in incident cases and control group,respectively.In the analysis of the associations between serum PFOA and diabetes risk factors,serum PFOA concentration was positively associated with Age(? = 0.0034,P = 0.03).Serum PFOA concentration of participants with higher education levels was significantly higher than those with lower education levels(1.38ng/m L vs.1.15ng/m L,P < 0.0001).Serum PFOA levels were positively correlated to BMI(? = 0.0089,P = 0.004),TC(? = 0.0316,P = 0.01),TG(? = 0.0368,P = 0.002),LDL(? = 0.0287,P = 0.04),and FPG(? = 0.0621,P = 0.0003).Participants with hyperlipidemia had higher serum BPA concentration than those without hyperlipidemia(1.35ng/m L vs.1.12ng/m L,P = 0.002).Serum PFOA concentration were significantly higher in subjects with FPG higher than 5.6mmol/L(1.33ng/m L vs.1.12ng/m L,P = 0.002)and 6.1mmol/L(1.36ng/m L vs.1.19ng/m L,P = 0.002)than in those with normal FPG.Serum PFOA levels were not significantly associated with gender,waist circumference,diet,and hypertension(all P > 0.05).Participants were classified into tertiles based on the serum PFOA concentration in the control group.The OR of tertile 3 was 1.56(95% CI,1.22-2.00,Ptrend < 0.0001),compared with tertile 1.After adjusted for traditional diabetes risk factors,the corresponding OR was 1.71(95%CI,1.11-2.65,Ptrend = 0.02).In stratified analysis,the association between serum PFOA and risk of type 2 diabetes was significant in participants with age < 65 years(OR = 1.58,95%CI,1.07-2.99,Ptrend = 0.02),males(OR = 1.76,95%CI,1.10-2.79,Ptrend = 0.02),middle school education or below(OR = 1.46,95%CI,1.01-2.10,Ptrend = 0.04),overweight and obesity(OR = 1.51,95%CI,1.00-2.28,Ptrend = 0.05),and central obesity(OR = 1.64,95%CI,1.07-2.50,Ptrend = 0.02),but no interaction of these diabetes risk factors and serum PFOA levels on type 2 diabetes risk was found(all Pinteraction > 0.05).FPG levels and BMI had significant mediating effects on the association between serum PFOA levels and the risk of type 2 diabetes(?= 0.132,0.031-0.243)and(?= 0.072,0.017-0.141),and 88% and 33.3% of the effects of serum PFOA levels on type 2 diabetes risk was mediated by FPG levels and BMI,respectively.No interaction of serum PFOA levels and GRS or w-GRS on type 2 diabetes was found(all Pinteraction > 0.05).In the analysis of the combined effect of GRS and serum PFOA levels on type 2 diabetes,participants were classified into tertiles based on the distribution of GRS and w-GRS,the ORs of the group with both highest GRS or w-GRS and serum PFOA concentration were 2.86(95%CI,1.60-5.13,P < 0.0001)and 3.19(95%CI,1.77-5.74,P < 0.0001),respectively,compared with those with both lowest GRS or w-GRS and serum PFOA concentration group.In both incident case and control groups,serum PFOA levels were not significantly associated with 5-years FPG levels changes(P > 0.05),and GRS or w-GRS did not modify the association between them.The median serum PFOS concentration was 11.98 and 11.82ng/m L in incident cases and controls,respectively.In the analysis of the associations between serum PFOS and diabetes risk factors,serum PFOS concentration in males was significantly higher than that in females(14.12ng/m L vs.10.58ng/m L,P = 0.01).Serum PFOS concentration was positively associated with BMI(? = 0.0163,P = 0.03).Serum PFOS levels were not associated with age,education,waist circumference,hypertension,hyperlipidemia,blood lipid,and FPG(all P > 0.05).Participants were classified into tertiles based on the serum PFOS concentration in the control group.The OR of tertile 3 was 1.12(95%CI,0.87-1.44,Ptrend = 0.45),compared with tertile 1.After adjusted for traditional diabetes risk factors,the corresponding OR was 1.12(95%CI,0.70-1.79,Ptrend = 0.61).In stratified analysis,no interactions of serum PFOS levels and diabetes risk factors on type 2 diabetes risk were detected.(all Pinteraction > 0.05).No interaction of serum PFOS levels and GRS or w-GRS on type 2 diabetes was found(all Pinteraction > 0.05).In the analysis of the combined effect of GRS and serum PFOS levels on type 2 diabetes,participants were classified into tertiles based on the distribution of GRS and w-GRS,the ORs of the group with both highest GRS or w-GRS and serum PFOS levels concentration were 1.42(95%CI,0.81-2.49,P = 0.22)and 1.60(95%CI,0.93-2.73,P = 0.09),respectively,compared with those with both lowest GRS or w-GRS and serum PFOS concentration group.In both incident case and control groups,serum PFOS levels was not significantly associated with 5-years FPG levels changes(P > 0.05),and GRS or w-GRS did not modify the association between them.Conclusion: Serum PFOA concentration was higher in older participants,with higher education levels,history of hyperlipidemia,higher BMI,blood lipid,and FPG.Serum PFOA was positively associated with the risk of type 2 diabetes,which was more significant in participants with age < 65 years,males,middle school education or below,centrally obesity,and overweight and obesity.No interaction of serum PFOA and these factors on type 2 diabetes was found.FPG and BMI had mediated 88% and 33.3% of the effects of serum PFOA on type 2 diabetes risk.Serum PFOA was not associated with 5-years FPG changes,and diabetes GRS did not modify the association between serum PFOA and type 2 diabetes risk and 5-years FPG changes.Individuals with both higher diabetes GRS and higher serum PFOA concentration had higher risk type 2 diabetes.Serum PFOS concentrations were higher in males with higher BMI.The present study did not find significant association between serum PFOS and the risk of type 2 diabetes and 5-years FPG changes.There was no modification and combined effects of GRS on the association of serum PFOS with the risk of type 2 diabetes and 5-years FPG changes.More prospective studies with larger sample size and more SNPs are warranted to confirm our findings in the future.Part III: Co-exposure to BPA,PFOA,PFOS,and the risk of type 2 diabetesObjective: To investigate the associations of co-exposure to serum BPA,PFOA,and PFOS with the risk of type 2 diabetes,and exploring their possible interaction and combined effects on type 2 diabetes risk.Methods: Participants' characteristics,the definition of type 2 diabetes,and serum BPA,PFOA,and PFOS measurements were the same as the first and second parts.Spearman correlation was used to estimate the correlation between serum BPA,PFOA,and PFOS.Conditional Logistic regression was used to estimate the associations between serum BPA,PFOA,and PFOS and type 2 diabetes risk in the case of simultaneous exposure of these three EDCs.The interaction and combined effects of serum BPA,PFOA,and PFOS with the risk of type 2 diabetes were assessed by unconditional Logistic regression model.Results: Serum PFOA and PFOS levels were significantly correlated(r = 0.58,0.579 and 0.578,respectively;all P <0.0001)in the incident case group,control group,and the total population,respectively.However,both serum PFOA and PFOS levels were not correlated with serum BPA levels(all P > 0.05).Participants were classified into tertiles based on the serum BPA,PFOA,and PFOS concentration in the control group.When serum BPA,PFOA,and PFOS were included in the same model,compared with tertile 1,the ORs of tertile 3 were 0.99(95%CI,0.80-1.24,Ptrend = 0.76),1.53(95%CI,1.17-1.98,Ptrend = 0.002),and 0.88(95%CI,0.68-1.14,Ptrend = 0.33)for serum BPA,PFOA,and PFOS,respectively.After adjustment for traditional type 2 diabetes risk factors,the corresponding ORs were 1.10(95%CI,0.81-1.48,Ptrend = 0.74),1.54(95%CI,1.08-2.19,Ptrend = 0.02),and 0.79(95%CI,0.56-1.13,Ptrend = 0.20),respectively.No interaction or combined effects of serum BPA,PFOA,and PFOS levels on the risk of type 2 diabetes was found(P > 0.05).Conclusion: Serum PFOA was positively associated with type 2 diabetes risk in the case of simultaneous exposure of BPA,PFOA,and PFOS.No interaction or combined effects of Serum BPA,PFOA,and PFOS on the risk of type 2 diabetes existed.more prospective studies with larger sample size are warranted to confirm our findings in the future. |
PDF Full Text Request