Associations Of 1-bromopropane Exposure And Genetic Factors With Fasting Plasma Glucose And Type 2 Diabetes

Type 2 diabetes(T2D)is a group of metabolic diseases characterized by hyperglycemia and is threating human health seriously.In recent decades,the prevalence of T2 D has grown dramatically in China,from 0.67% in 1980 to 9.7% in 2010,and has caused a huge disease burden.Recent studies have suggested environmental pollutants to be important risk factors of T2 D in addition to the traditional risk factors such as overweight and unhealthy lifestyles.Moreover,considering that genetic factors can play an important role in the development of T2 D,the interaction between environmental factors and genetic factors on T2 D has gradually become a research hotspot.1-bromopropane(1-BP)is a compound formed from the substitution of the 1-position hydrogen atom of propane by bromine atom.1-BP is a kind of volatile colorless liquid with a slightly irritating odor under room temperature.With the advantages with Freon and harmlessness to ozone layer,1-BP has been used as an alternative for Freon in the synthesis of industrial products and daily necessities including cleaning agent for metal and precision electronic components,degreaser,dry cleaner,spray adhesives,insecticides,flavors,pharmaceuticals,and aerosol spray.The rapid growth of 1-BP usage has caused a widely exposure to occupational and general population,and has raised great concerns on human health.Previous studies have suggested a neurotoxicity of 1-BP,and the International Agency for Research on Cancer has classified 1-BP into possibly carcinogenic to humans(Group 2B).Although the results from animal experiment researches have showed that 1-BP could damage metabolic organs like liver and epidemiological studies have suggested a positive association between 1-BP exposure and blood glucose levels,the results have not been evaluated in a cohort with large sample size.Above all,it is imperative to investigate the association between 1-BP exposure and the risk of T2 D,as well as the underlying mechanisms.1-BP exposure can occur through inhalation and dermal contact.Except for exhalation of parent material or CO2,1-BP is mainly excreted through urine in the form of oxygenated and conjugated metabolites.As a metabolite of 1-BP,urinary N-Acetyl-S-(n-propyl)-L-cysteine(BPMA)has been recognized as a specific and sensitive biomarker of 1-BP internal exposure levels.Besides,previous studies have revealed that 1-BP would exhaust glutathione during the metabolic process and lead to a high level of oxidative stress and consequent inflammation.The present study investigated the associations of 1-BP exposure with fasting plasma glucose(FPG)and T2 D in community population,and evaluated the potential mediating roles of oxidative stress and inflammation.Furtherly,the interactions of genetic factors and 1-BP exposure with FPG and T2 D were explored.We developed the present study with urban Chinese participants from the Wuhan-Zhuhai cohort.Urinary BPMA levels of the baseline and the follow-up population were determined.First,the distribution of urinary BPMA levels was analyzed and the potential influence factors were explored.Second,the cross-sectional and longitudinal associations of 1-BP exposure with FPG and T2 D were estimated.Third,urinary 8-hydroxy-deoxy-guanosine(8-OHd G),urinary 8-iso-prostaglandin-F2α(8-iso-PGF2α),and serum C-reactive protein(CRP)were determined as biomarkers of DNA oxidative damage,lipid peroxidation,and systematic inflammation,respectively.The mediating roles of oxidative stress and inflammation in the association between 1-BP exposure and FPG were investigated to explore the possible mechanisms underlying.Finally,the T2D-related polygenetic risk scores(PRS)based on the results of genome-wide association analysis(GWAS)were calculated and the interactions between PRS and 1-BP exposure on FPG and T2 D were evaluated.The present study protocol was approved by the Ethics and Human Subject Committee of Tongji Medical College,Huazhong University of Science and Technology and all participants enrolled in this study provided written informed consent.A robust quality control system has been established to ensure the accuracy and reliability of the data used in analyses.Our study including the following four parts:Part One: The distribution of 1-BP internal exposure levels among community population and potential influence factorsObjectives: To evaluate the levels and distribution of 1-BP internal exposure in WuhanZhuhai community population,and to explore the possible influence factors.Methods: The participants came from the baseline of the Wuhan-Zhuhai cohort during 2011 to 2012.Detailed information regarding demographic characteristics,lifestyles,living condition,disease history,and medication history was obtained from face-to-face investigations.Urinary BPMA levels was determined with early morning urine sapmles using an ultra-high performance liquid chromatography system coupled with electrospray ionization tandem mass spectrometry(UHPLC-ESI/MSMS).Univariate models with creatinine corrected BPMA as the dependent variable were used to compare the differences of BPMA across subgroups.Multivariable analyses were further conducted with the adjustion of potential confounders including gender,age,body mass index(BMI),cigarette smoking status,alcohol drinking status,self-cooking,ventilation devices in kitchen,traffic exposure time,community,and the consumption frequency of vegetable and fruit,meat,aquatic products,and egg and milk.All p-values were two sided with a significant level at 0.05,and all data were analyzed with SAS 9.4(SAS Institute Inc.Cary,NC,USA).Results: The present study included 3749 participants from the baseline of WuhanZhuhai cohort with a mean age of 52.72±12.84 years.The limit of detection of urinary BPMA levels was 3 ng/m L and the detectable rate was 99.5%.The geometric mean of urinary BPMA levels was 10.60 ng/m L and the geometric mean of creatininecorrected urinary BPMA levels was 0.94 μg/mmol Cr.When compared with previous studies,the urinary BPMA levels in the present study was lower than that among occupational workers(37.29-4644.30 μg/mmol Cr),and higher than that among American general population(0.38-0.59 μg/mmol Cr).The results of univariate models showed that advanced age,lower aquatic products consumption frequency,self-cooking,and no ventilation device in kitchen were associated with the elevated urinary BPMA levels(P<0.05).In multivariable models,the olders(>55 years old),females,and cigarette smokers showed higher urinary BPMA levels than their colleagues(P<0.05).Further analyses showed that cigarette smoking status did not change urinary BPMA levels in males or females(P>0.05),and gender did not change urinary BPMA levels in smokers or nonsmokers(P>0.05).Conclusions: The urinary BPMA levels in the present study was lower than that among occupational workers,and higher than that among American general population.Advanced age may lead to an elevated urinary BPMA levels.Part Two: The cross-sectional and longitudinal associations of 1-BP exposure with FPG and T2DObjectives: To investigate the cross-sectional and longitudinal associations of 1-BP exposure with FPG and T2 D.Methods: Participants with complete information on FPG and T2 D from part one were included and were followed-up every 3 years.T2 D was diagnosed among the individuals with FPG≥7.0 mmol/L,or antidiabetes drug usage,or clinically diagnosed T2 D.The urinary BPMA levels were logarithmically transformed due to its right-skewed distribution.In cross-sectional analyses,general linear models were used to analyse the association between BPMA and FPG,and logistic regression models were used to calculate the odds ratios(ORs)and 95% confidence interval(CI)for prevalent T2 D associated with 1-BP exposure.In longitudinal analyses,participants were divided into high exposure group and low exposure group according to the median levels of BPMA in each research year.And then participants were divided into three groups,i.e.persistent low exposure(low exposure in both baseline and follow-up),inconsistent exposure(different exposure groups between baseline and follow-up),and persistent high exposure(high exposure in both baseline and follow-up).General linear models with the difference values of FPG between follow-up and baseline as the dependent variable and the BPMA group as the independent variable were used to analyse the association between BPMA and the change of FPG.Cox regression models were used to calculate the hazard ratios(HRs)for incident T2 D associated with BPMA.Longitudinal analyses were conducted in a 3-year interval and a 6-year interval,respectively.All models were adjusted for potential confounders including gender,age,BMI,cigarette smoking status,alcohol drinking status,physical exercise,ventilation devices in kitchen,consumption frequency of aquatic products,family history of diabetes,antidiabetes drug usage,and community.All pvalues were two sided with a significant level at 0.05,and all data were analyzed with SAS 9.4(SAS Institute Inc.Cary,NC,USA).Results: In the cross-sectional analyses,the results from general linear models showed that each 1-unit increase in the log-transformed BPMA values was associated with a 0.091 mmol/L(95% CI 0.038 to 0.144)increase in FPG levels.In comparison with participants in the first quartile of BPMA,the adjusted β(95% CI)for FPG levels among the second,third,and fourth BPMA quartile were 0.004(-0.138 to 0.146),0.076(-0.066 to 0.219),and 0.271(0.126 to 0.415)mmol/L,respectively.Gender and cigarette smoking status statistically significantly modified the association between BPMA and FPG,with stonger associations were detected among males or smokers.No modification effect by other covariates were detected.The results from logistic regression models showed that each 1-unit increase in the log-transformed BPMA values was associated with a 34.8%(OR 1.348;95% CI 1.187-1.531)increase in the prevalent risk of T2 D.Compared with participants in the first quartile of BPMA,the adjusted OR(95% CI)for T2 D in the second,third,and fourth BPMA quartile were 0.949(0.627-1.436),1.542(1.058-2.246),and 2.250(1.571-3.224),respectively.The results from interaction analyses demonstrated an additive interaction between BPMA and age on the prevalent risk of T2D(RERI 1.881,95% CI 0.657 to 3.106).In the longitudinal analyses,the results from general linear models showed that the adjusted β(95% CI)for the change of FPG in the persistent high exposure groups were 0.262(0.088 to 0.436)mmol/L in a 3-year scale and 0.237(0.068 to 0.406)mmol/L in a 6-year scale,respectively,when compared with the persistent low exposure groups.The results from Cox regression models showed that the adjusted HR(95% CI)for T2 D in the persistent high exposure groups were 1.951(1.084-3.512)in the 3-year scale and 1.654(1.003-2.727)in the 6-year scale,respectively,when compared with the persistent low exposure groups.Conclusions: In cross-sectional study,urinary BPMA levels were significantly positively associated with FPG levels and the prevalent risk of T2 D.The results from longitudinal study suggested that urinary BPMA levels were significantly positively associated with the change of FPG,and the incident risk of T2 D was significantly increased for participants in persisitent high exposure groups compared with participants in persisitent low exposure groups.Part Three: The mediating effects of oxidative stress and systematic inflammation in the associations of 1-BP exposure with FPG and T2DObjectives: To evaluate the mediation effects of oxidative stress and inflammation in the associations of urinary BPMA levels with FPG levels and the prevalent risk of T2 D in community population.Methods: Urinary 8-OHdG,urinary 8-iso-PGF2α,and plasma CRP levels were determined in the baseline of the Wuhan-Zhuhai cohort.General linear models were used to investigate the associations of BPMA with 8-OHd G,8-iso-PGF2α,and CRP,and the associations of 8-OHd G,8-iso-PGF2α,and CRP with FPG,respectively.Logistic regression models were used to calculate the ORs for T2 D associated with 8-OHd G,8-iso-PGF2α,and CRP,respectively.Mediation effect models were used to evaluate the mediating roles of 8-OHd G,8-iso-PGF2α,and CRP in the associations of BPMA with FPG and T2 D,respectively.Models were adjusted for potential confounders including gender,age,BMI,cigarette smoking status,alcohol drinking status,physical exercise,ventilation devices in kitchen,consumption frequency of aquatic products,family history of diabetes,antidiabetes drug usage,and community.All p-values were two sided with a significant level at 0.05,and all data were analyzed with SAS 9.4(SAS Institute Inc.Cary,NC,USA).Results: Each 1-unit increase in the log-transformed BPMA values was associated with a 0.122(95% CI 0.074 to 0.170)increase in log-transformed 8-OHd G values,and a 0.280(95% CI 0.255 to 0.305)increase in log-transformed 8-iso-PGF2α values.Compared with participants in the first quartile of BPMA,the adjusted β(95% CI)for 8-OHd G in the second,third,and fourth BPMA quartiles were 0.097(-0.030 to 0.223),0.177(0.049 to 0.304),and 0.263(0.134 to 0.392),respectively;and for 8-iso-PGF2α were 0.278(0.209 to 0.346),0.461(0.392 to 0.529),and 0.639(0.569 to 0.708),respectively.The association between BPMA and CRP was statistically insignificant(P>0.05).After adjusting for potential confounders,each 1-unit increase in the logtransformed 8-iso-PGF2α values was associated with a 0.075 mmol/L increase in FPG levels.Compared with participants in the first quartile of 8-iso-PGF2α,the adjusted β(95% CI)for FPG in the second,third,and fourth 8-iso-PGF2α quartiles were-0.062(-0.205 to 0.081),0.112(-0.035 to 0.259),and 0.172(0.017 to 0.328)mmol/L,respectively.The relationships of 8-OHd G or CRP with FPG were statistically insignificant(P>0.05).In univariate models,each 1-unit increase in the log-transformed 8-OHd G,8-isoPGF2α,and CRP values were associated with a10.9%(OR 1.109;95% CI 1.005-1.224),20.6%(OR 1.206;95% CI 1.051-1.384),and 21.1%(OR 1.211;95% CI 1.114-1.317)increase in the prevalent risk of T2 D,respectively.In multivariate modles,the associations of 8-OHd G,8-iso-PGF2α,or CRP with the prevalent risk of T2 D were statistically insignificant(P>0.05).The results from mediation effectanalyses showed that 8-iso-PGF2αmediated the association between BPMA and FPG with the mediation proportion of 30.26%;while no mediation effects were found for 8-OHd G or CRP.Moreover,no mediation roles were found in the association between BPMA and the prevalent risk of T2 D for 8-OHd G,8-iso-PGF2α,or CRP.Conclusions: Urinary BPMA levels was associated with DNA damage and lipid peroxidation,and lipid peroxidation may be one of the potential mechanisms underlying the association between urinary BPMA levels and FPG levels.Part Four: The interactions between genetic factors and 1-BP exposure on FPG and T2DObjectives: To investigated the associations of genetic factors with FPG and T2 D,and evaluate the interactions between genetic factors and 1-BP exposure on FPG and T2 D.Methods: DNA extraction and genotyping were conducted for participants using Illumina Human Omni Zhong Hua-8 chips in the present study.Considering no public database contained GWAS genotyping information from Chinese population,we chose Japanese population which is similar to Chinese population to aquire the T2 D related SNPs and corresponding β value for the incident risk of T2D(from Japan Biobank).The PRS for each individual in the present study was calculated from the sum of the product of the risk allele number and the corresponding β value for each SNP.Mixed linear models and general linear models were used to investigate the cross-sectional and longitudinal associations between PRS and FPG,respectively.Logistic regression models and Cox regression models were used to investigate the cross-sectional and longitudinal associations between PRS and T2 D,respectively.Participants were divided into three groups according to the tertiles of PRS,and the cross-sectional and longitudinal associations of BPMA with FPG and T2 D were evaluated in each PRS group.Furtherly,we divided participants into nine groups according to the tertiles of BPMA and the tertiles of PRS,and the joint effect between PRS and BPMA on FPG and T2 D were evaluated.All models were adjusted for potential confounders including gender,age,BMI,cigarette smoking status,alcohol drinking status,physical exercise,ventilation devices in kitchen,consumption frequency of aquatic products,and the first five principal components of ancestry.Considering that PRS contained genetic factors and the first five principal components represented the population structure differences,family history of diabetes and community were not adjusted.All p-values were two sided with a significant level at 0.05,and all data were analyzed with EAGLE2,Minimac4,PLINK(v1.90b6.2),SAS 9.4(SAS Institute Inc.Cary,NC,USA)and R 3.6.3(R Core Team).Results: In the cross-sectional analyses,each 1-unit increase in PRS was associated with a 0.240 mmol/L(95% CI 0.147 to 0.333)increase in FPG levels.Compared with participants in the lower tertile of PRS,the adjusted β(95% CI)for FPG the middle and higher PRS tertiles were 0.103(-0.011 to 0.217)and 0.273(0.159 to 0.388),respectively.Each 1-unit increase in PRS was associated with a 126.7%(OR 2.267;95% CI 1.711-3.004)increase in the prevalent risk of T2 D.Compared with participants in lower tertile of PRS,the adjusted OR(95% CI)for T2 D in the middle and higher PRS tertiles were 1.604(1.102-2.335)and 2.378(1.669-3.388),respectively.In the cross-sectional analyses,for participants in the lower,middle,and higher PRS tertiles,each 1-unit increase in the log-transformed BPMA values was associated with a 0.085 mmol/L,0.159 mmol/L,and 0.208 mmol/L increase in FPG levels,respectively;and a 19.3%,32.3%,and 40.2% increase in the prevalent risk of T2 D,respectively.The results of joint effect analyses showed that the adjusted β(95% CI)for FPG and the adjusted OR(95% CI)for T2 D in the high PRS and high BPMA group were 0.669(0.494 to 0.844)and 3.027(1.704-5.376),respectively,when compared with participants in the low PRS and low BPMA group.There was a statistically significant interaction between PRS and BPMA on FPG(P for interaction <0.001)while no interaction was found for T2 D.In the 3-year longitudinal analyses,each 1-unit increase in PRS was associated with a 0.224 mmol/L(95% CI 0.099 to 0.350)increase in the change of FPG,and a 98.2%(HR 1.982;95% CI 1.333-2.948)increase in the incident risk of T2 D.In the 6-year longitudinal analyses,each 1-unit increase in PRS was associated with a 0.124 mmol/L(95% CI-0.001 to 0.249)increase in the change of FPG,and a 69.1%(HR 1.691;95% CI 1.171-2.443)increase in the incident risk of T2 D.In the longitudinal analyses,the statistic significant results of association of 1-BP exposure with FPG and T2 D were acquired only in higher PRS tertiles.In higher PRS tertiles,compared with the BPMA persistent low exposure groups,the adjusted β(95% CI)for the change of FPG in the BPMA persistent high exposure groups were 0.412(0.043 to 0.782)in the 3-year scale and 0.508(0.144 to 0.872)in the 6-year scale,respectively.Compared with the BPMA persistent low exposure groups,the adjusted HR(95% CI)for T2 D in the BPMA persistent high exposure groups were 2.289(0.879-5.959)in the 3-year scale and 2.761(1.201-6.346)in the 6-year scale,respectively.Conclusions: There was a statistically significant additive interaction between PRS and 1-BP exposure on FPG levels,while no interaction was detected for T2 D.