| Purpose: There are few studies on the relationship between perfluoroalkyl substances(PFASs)exposure and non-alcoholic fatty liver disease(NAFLD).This study intends to use the US National Health and Nutrition Examination Survey(NHANES)database to explore the relationship between perfluoroalkyl substances and non-alcoholic fatty liver disease,and the results can provide scientific basis for the prevention of non-alcoholic fatty liver disease and the exposure management of perfluoroalkyl substances.Methods: We selected data from seven consecutive cycles of the NHANES database(2005-2006,2007-2008,2009-2010,2011-2012,2013-214,2015-2016,2017-2018)to screen adults aged 20 years and older with complete data and no competitive liver disease.Perfluorooctanoic acid(PFOA),perfluorooctane sulfonic acid(PFOS),fluorohexane sulfonic acid(PFHx S)and perfluorofluorononanoic acid(PFNA)four PFASs were selected as independent variables.NAFLD was defined by the US non-alcoholic fatty liver index(US FLI).(1)t test and χ2 test were used to compare the basic characteristics of non-alcoholic fatty liver disease between those who did not have the NAFLD and those who did.(2)Poisson regression analysis was conducted to evaluate the association between the four PFASs exposures and the risk of NAFLD,and the relative risks(RR)and 95% confidence interval(CI)were calculated.Covariates were age,sex,race,household income to poverty ratio,occupation,education,body mass index,waist circumference,physical activity,smoking status,insulin resistance,comorbidities,hypertension,hyperlipidemia,energy intake and survey period.Besides,women also adjusted for menstrual status.Firstly,four PFASs were divided into Q1,Q2,Q3 and Q4 groups by quartile method.The lowest dose group Q1 was used as the reference group to calculate the RRs and 95% CIs of Q2,Q3 and Q4.Secondly,the concentrations of ln-PFASs after logarithmic transformation of four PFASs were used as continuous quantitative variables to enter Poisson regression.Then,principal component analysis was performed on the four PFASs,and a new variable "comprehensive PFASs" was generated.The new variable entered the Poisson regression as a continuous quantitative variable to evaluate the influence of the combined action of the four PFASs on the outcome.In addition,stratified analysis was conducted according to sex,age and body mass index.(3)Restricted cubic spline models were used to assess the dose-response relationship between PFASs exposure and risk of NAFLD.(4)Multiplication and addition interaction models were used to analyze the interaction between PFASs exposure and NAFLD risk.Results: According to inclusion and exclusion criteria,a total of 3,464 subjects were included in this study,with an average age of 53.133±17.952 years.Among them,1,612 cases were male,accounting for 46.54%;There were 1,852 cases of female patients,accounting for 53.46%.1.Prevalence of NAFLD In this study,there were 1200 cases of non-alcoholic fatty liver disease,the total prevalence rate was 34.64%,including 637 cases of male disease,the prevalence rate was39.52%;There were 563 cases in females with a prevalence of 30.40%,among which270 cases in premenopausal females with a prevalence of 26.01% and 293 cases in postmenopausal females with a prevalence of 36.00%.2.Association between four PFASs exposures and NAFLD risk(1)Association between four PFASs exposures and NAFLD in the general population Compared with the respective Q1 dose groups,there was no statistically significant association between Q2,Q3,Q4 dose groups of PFOA,PFOS,PFHx S and PFNA exposure and risk of NAFLD(p>0.05),nor was there statistically significant association between "combined PFASs" and risks of NAFLD(p>0.05).(2)The association between four PFASs exposures and the risk of NAFLD in different sexes After adjusting for covariates,there was no statistical significance in Q2,Q3 and Q4 dose groups of PFOA,PFOS,PFHx S and PFNA in male population compared with the lowest dose group of Q1(p>0.05).In the female population,there was a statistically significant association between exposure to PFOA,Q3 and Q4 dose groups and the risk of NAFLD,with RRs(95%CIs)of 1.640(1.117,2.407)and 1.867(1.210,2.879),respectively,and the trend test was statistically significant(p-trend<0.05).There was a statistically significant association between Q4 dose of PFOS exposure and risk of NAFLD(p<0.05).There was a statistically significant association between exposure to PFNA in Q3 and Q4 dose groups and the risk of NAFLD,with RRs(95%CIs)of 1.983(1.368,2.874)and2.000(1.242,3.221),respectively,and the trend test was statistically significant(p-trends<0.05).The logarithm-transformed serum concentrations of the four PFASs were introduced into regression using continuous variable ln-PFASs.The RR(95%CI)between ln-PFOS and NAFLD was 0.878(0.778,0.991)in the male population,and in the female population,The RRs(95% CIs)of ln-PFOA and ln-PFNA were 1.354(1.110,1.651)and 1.244(1.007,1.537),respectively.The association between "combined PFASs" and the risk of NAFLD was statistically significant(p<0.05),with a RR(95%CI)of 1.137(1.003,1.289)in women.(3)Association between exposure to four PFASs and risk of NAFLD in women with different postmenopausal status After adjusting for covariates,the Q3 and Q4 dose groups of PFOA,PFOS and PFNA were significantly associated with the risk of NAFLD in premenopausal women compared with the lowest dose group of Q1(p<0.05).The RRs(95%CIs)of PFOA were1.831(1.071,3.130)and 2.307(1.204,4.419),and showed statistical significance in trend test(p-trends <0.05).In postmenopausal women,there was a statistical difference between exposure to PFOA in Q3 and Q4 dose groups and risk of NAFLD,with RRs(95%CIs)being 2.148(1.257,3.671)and 2.331(1.207,4.504),respectively,and trend test was statistically significant(p-trend <0.05).There was a statistical difference between Q2 and Q3 exposure to PFHx S and the risk of NAFLD,with RRs(95%CIs)of1.868(1.158,3.011)and 2.131(1.075,4.226),respectively.The log-converted serum concentrations of the four PFASs were introduced by regression as a continuous variable ln-PFASs,and the RRs(95% CIs)of ln-PFOA and ln-PFNA in premenopausal women were 1.422(1.031,1.962)and 1.395(1.021,1.906),respectively.In postmenopausal women,the RRs(95% CIs)of ln-PFOA and ln-PFHx S were 1.692(1.241,2.307)and 1.390(1.115,1.732),respectively.The association between "combined PFASs" and risk of NAFLD was statistically significant in premenopausal women,with a RR(95%CI)of 1.232(1.047,1.450).(4)Association between exposure to four PFASs and risk of NAFLD in different age groups In three age groups(20-39,40-59,and ≥60 years),only 60 years and older age groups were found to have a statistically significant association between PFNA exposure and risk of nonalcoholic fatty liver disease,with an RR(95%CI)of 2.292(1.122,4.683).(5)Association between exposure to four PFASs and risk of NAFLD in different BMI groups The association between ln-PFOA and "combined PFASs" exposure and risk of NAFLD was statistically significant in the normal/low weight group(p<0.05),with RRs(95%CIs)of 0.325(0.119,0.893)and 0.526(0.287,0.965),respectively.The association between ln-PFOA and "combined PFASs" exposure and risk of NAFLD was statistically significant in the normal/underweight population(p<0.05).The association between ln-PFNA and the risk of NAFLD was statistically significant in the obese group,with a RR(95%CI)of 1.303(1.012,1.677).3.Dose-response relationship between exposure to four PFASs and risk of NAFLD The results of restricted cubic spline showed that the risk of NAFLD increased monotonically with the increase of PFOA concentration over 2.89 ng/ml and PFNA concentration over 1.10 ng/ml in female population.In premenopausal women,when the concentration of PFOA was greater than 3.47 ng/ml,and when the concentration of PFNA was greater than 2.10 ng/ml,the risk of NAFLD increased monotonically with the increase of concentration.In postmenopausal women,the risk of NAFLD increased first and then decreased with the increase of PFOA concentration,and the risk of NAFLD was the highest when PFOA concentration was 3.57ng/ml.With the increase of PFHx S concentration,the risk of NAFLD increased first and then decreased,and the highest risk was found when the PFHXS concentration was 3.1ng/ml.4.Analysis of the interaction between four PFASs exposure,gender and smoking on NAFLD The interaction analysis of four PFASs substances with sex and smoking respectively showed that the multiplicative interaction of PFOA,PFOS and PFHx S with smoking was statistically significant,and the RRs(95%CIs)were 1.127(1.007,1.262)and 1.113(1.009,1.227)and 1.095(1.004,1.194),respectively.Conclusions: Exposure to four PFASs alone or in combination was associated with risk of NAFLD,and the associations were different between sexes,age groups,postmenopausal status,and BMI grades,and had a multiplicative interaction with smoking. |
PDF Full Text Request