Positive Association Between Blood Trihalomethane Concentrations and Non-Alcoholic Fatty Liver Disease: A Cross-Sectional Study (NHANES)

doi:10.21203/rs.3.rs-5034617/v1

Download PDF

Research Article

Positive Association Between Blood Trihalomethane Concentrations and Non-Alcoholic Fatty Liver Disease: A Cross-Sectional Study (NHANES)

https://doi.org/10.21203/rs.3.rs-5034617/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background: Trihalomethanes (THMs) is a common byproduct of disinfection that has been shown to be hepatotoxic. However, the relationship between THMs and non-alcoholic fatty liver disease (NAFLD) remains unclear.

Methods: This study selected 9475 adults from the National Health and Nutrition Examination Survey (NHANES) from 1999 to 2012, and the concentrations of various THMs including chloroform (TCM), bromodichloromethane (BDCM), dibromochloromethane (DBCM), and bromoform (TBM) in their blood were analyzed. NAFLD was identified based on the levels of the fatty liver index (FLI), Alanine Aminotransferase (ALT), and Aspartate Aminotransferase (AST).

Results: In the multiple logistic regression model, we found TBM, Br-THM and TTHM concentrations were significantly positively correlated with NAFLD, The odds ratios (ORs) were 1.27 (95% CI 1.07-1.50), 1.19 (95% CI 1.01-1.40), and 1.27 (95% CI 1.07-1.52), respectively, indicating the risk of NAFLD was on the rise with the increase of these THM concentrations. Although the ORs of blood TCM and Cl-THM concentrations were not significant, there was a trend suggesting an increased risk of NAFLD with the increase of their concentrations.

Conclusion: This study suggested that THMs exposure is associated with NAFLD in the USA population, and more prospective studies are still needed to confirm this finding and elucidate the underlying mechanisms in the future.

Trihalomethanes

Non-alcoholic fatty liver disease

Byproduct

Correlation

Non-alcoholic fatty liver disease (NAFLD) is a common liver disease, which is characterized by fat deposits in the liver in the absence of alcohol abuse, and can even escalate to non-alcoholic steatohepatitis (NASH), cirrhosis, and hepatocellular carcinoma [1–3]. NAFLD has become a major global health issue, and its prevalence is on the rise [4]. Epidemiological studies indicate that the prevalence of NAFLD in the global adult population is approximate 25%, and even more than 30% in some developed countries and regions [5–7]. The pathogenesis of NAFLD is complex, mainly involving many factors such as hepatic lipid metabolism disorder, insulin resistance, oxidative stress, and inflammatory response [8, 9]. Metabolic syndrome (MetS) and type 2 diabetes mellitus (T2DM) are significant risk factors for NAFLD [10, 11]. Additionally, genetic predispositions, intestinal microecological imbalances, and exposure to environmental pollutants also contribute to the occurrence and development of NAFLD [12–15]. Trihalomethanes (THM) is a common disinfection by-product of drinking water, which mainly includes chloroform (CHCl₃), bromodichloromethane (CHBrCl₂), dibromochloromethane (CHBr₂Cl), and bromoform (CHBr₃), accounting for 66% of disinfection by-products [16]. The formation of THM is mainly due to the halogenation of hypochlorite (HOCl) and natural organic matter (NOM) during water disinfection. Due to its volatility, THM can be absorbed or inhaled through drinking water, swimming, showering, bathing, and cooking [17]. Despite drinking water disinfection has health benefits, studies have shown that long-term exposure to high level of THM may have adverse effects on the liver and kidneys, and even have potential carcinogenic risk [18, 19]. Animal studies have shown that exposure to bromodichloromethane (BDCM) can cause liver damage in obese mice, showing early signs of steatohepatitis [20, 21]. A case-control study by Makris et al. found a significant correlation between urinary THM concentration and serum alanine aminotransferase (ALT) level, and highlighted that exposure to environmental Br-THM can affect liver function, suggesting that Br-THM may be a potential liver toxin [22]. In addition, similar conclusion was also found by Burch et al [23]. Together, these studies indicated that there is a potential association between THM exposure and liver damage. However, the systematic research on the association between THM and NAFLD is limited. The further investigation of the potential connection between them will not only help shed light on the pathogenesis of NAFLD, but also provide a new perspective for the prevention and control of this disease.

2.1. Study population

The National Health and Nutrition Examination Survey (NHANES) is a vital research program whose goal is to assess the health and nutritional status of adults and children in the United States (USA). This program uniquely combines interviews and physical examinations to evaluate a nationally representative sample of approximately 5,000 individuals each year and publish collected findings biennially [24]. The detailed methods for data collection have been previously reported [25]. NHANES program was approved by the National Center for Health Statistics (NCHS) Ethics Review Board, and all participants signed written informed consent. Due to the too low detection rate of THM after 2013, our analysis focused on data of seven survey cycles from 1999 to 2012. Of the 71,916 participants, there were 38,024 adults (≥ 20 years of age), of which 33,606 participants could be identified as NAFLD. After excluding participants with other causes of liver disease (n = 5325), excessive alcohol consumption (n = 1578), covariator absence (n = 1578), and absence of valid THM information, the left in analysis included 9,475 participants (Fig. 1).

2.2. Blood THM measurements

During each examination cycle, additional peripheral venous blood samples are collected from selected participants specifically for volatile organic compound (VOC) studies. These samples are immediately and securely stored and transported to the laboratory [26]. The concentrations of chloroform (TCM), bromodichloromethane (BDCM), dibromochloromethane (DBCM), and bromoform (TBM) in whole blood were measured using solid-phase microextraction (SPME), gas chromatography (GC), and isotope dilution mass spectrometry (IDMS) [27]. For analytes whose concentrations were below the lower limit of detection (LOD), the values were interpolated as LOD/√2 divided by the LOD [28]. In addition to measuring the above four single THMs, we also calculated the concentrations of brominated THM (Br-THM and sum of BDCM, DBCM, and TBM), chlorinated THM (Cl-THM and sum of TCM, BDCM, and DBCM), and total THMs (TTHMs and sum of the four THMs).

2.3 Definition of NAFLD

NAFLD was defined in the following two conditions: (1) The fatty liver index (FLI) was calculated using the following formula, and its value ranges from 0 to 100. NAFLD was diagnosed when the FLI value was ≥ 60 [29].

$$\:FLI=\left(\frac{{e}^{0.953\cdot\:\text{ln}\left(TG\right)+0.139\cdot\:BMI+0.718\cdot\:\text{ln}\left(GGT\right)+0.053\cdot\:\text{w}\text{a}\text{i}\text{s}\text{t}\:\text{c}\text{i}\text{r}\text{c}\text{u}\text{m}\text{f}\text{e}\text{r}\text{e}\text{n}\text{c}\text{e}-15.745}}{1+{e}^{0.953\cdot\:\text{ln}\left(TG\right)+0.139\cdot\:BMI+0.718\cdot\:\text{ln}\left(GGT\right)+0.053\cdot\:\text{w}\text{a}\text{i}\text{s}\text{t}\:\text{c}\text{i}\text{r}\text{c}\text{u}\text{m}\text{f}\text{e}\text{r}\text{e}\text{n}\text{c}\text{e}-15.745}}\right)\times\:100$$

(2) The elevated liver enzymes were also diagnostic criteria: AST > 31 U/L or ALT > 31 U/L in females, and AST > 37 U/L or ALT > 40 U/L in males [30, 31]. NAFLD was diagnosed if any of the above criteria were met and no of the following conditions are present: positive serum hepatitis B surface antigen, positive hepatitis C antibody or HCV RNA, or heavy alcohol consumption (≥ 1 alcoholic drink/day for females or ≥ 2 alcoholic drink/day for males) [32, 33].

2.4. Assessment of Covariates

This study included age, gender, race, education, body mass index (BMI), smoking, physical activity, diabetes status, and survey cycles as covariates. The races were classified into five categories such as Mexican American, other Hispanic, non-Hispanic white, non-Hispanic black, and other/multiracial. The level of education was categorized as below high school, high school, and some college or more. Smoking status was divided into three groups such as never-smokers, former smokers, and current smokers, based on whether they had smoked fewer than 100 cigarettes in their lifetime and whether they had quit. Diabetes is defined based on the following criteria: (1) having been diagnosed with diabetes or taking insulin (2) hemoglobin A1C concentration > 6.5%, (3) fasting blood glucose level ≥ 126 mg/dL [34].

2.5. Statistical Analysis

The population was divided into NAFLD and non-NAFLD groups based on pre-determined criteria. Descriptive statistical methods were used to describe the statistical characteristics of the population. In the analyses, we considered the study sampling scheme, stratification, and subsample weights to ensure the results are representative. The statistical differences between the two groups were calculated using t-test for continuous variables and the Chi-square test for categorical variables. Multiple logistic regression analysis was conducted to assess the odds ratios (OR) and 95% confidence intervals (CI) for the association between each THM and NAFLD. The population was assigned to the quartiles of TCM, Br-THM, Cl-THM, and TTHM. Due to low detection rates, BDCM was grouped by tripartites, DBCM is < 50 th, 50 th-75 th, and > 75 th percentiles, and TBM is < 75 th, 75 th − 87.5 th, and > 87.5 th percentiles [35]. The p for trend was calculated based on the quartiles of each category as a continuous numeric variables. In model 1, no covariates were adjusted. In model 2, adjustments were made for age, gender, ethnicity, and survey cycles. In model 3, was further adjusted for age, gender, race, survey cycles, education level, smoking status, physical activity status, and BMI. In model 4, diabetes was added as a covariate [36]. For TTHM, we stratified and analyzed by age (20–40, 41–60, > 60), sex (male, female), race (Mexican American, other Hispanic, non-Hispanic white, non-Hispanic black, other/multiracial), BMI (below/normal weight, overweight, obese), smoking status (never, former, current), alcohol consumption (yes, no), diabetes status (yes, no), and physical activity level (heavy, moderate, light). The multiplicative interactions between TTHM and these stratified variables was judged by the likelihood ratio test [16]. At the same time, we also conducted four sensitivity analyses. Firstly, we excluded participants with missing BMI (n = 107) to determine the effect of interpolation on the findings. Secondly, in order to minimize the effects of peak exposure, we excluded participants who had used swimming pools, hot tubs, or steam rooms in the past 72 h (n = 260). Thirdly, we included the time interval since the last bath or shower (≤ 2, 3–6, 7–14, or > 14 h) in the multivariate model. Finally, we adjusted for the total concentrations of all other THMs to analyze the common exposure effects by including them as covariates in the multivariate model.

3.1. Population Characteristics.

Of the 9,475 participants included in the study, there were 1,638 NAFLD patients and 7,837 non-NAFLD patients. Compared to non-NAFLD participants, there was a tendency for NAFLD patients to be younger (mean age 45.2 years, 95% CI 44.4–45.9, versus 47.7 years, 95% CI 47.3–48.1), more likely to be male (50.7% versus 45.0%), have higher BMIs (mean 32.9 kg/m², 95% CI 32.5–33.4 versus 29.2 kg/m², 95% CI 29.1–29.3), also more likely to be Mexican American (25.4% versus 18.1%), engage in light physical activity more frequently (52.5% versus 48.9%), and more likely to have diabetes (21.7% versus 14.5%). The remaining population characteristics showed no significant differences between the two groups (Table 1).

Table 1

Survey-weighted Characteristics of Study Participants in NHANES 1999 − 2012 ^{a, b}.
Characteristics^c	All (n = 9475)	NAFLD		p Value
Characteristics^c	All (n = 9475)	No (n = 7837)	Yes (n = 1638)	p Value
Age (years)	47.2(46.9,47.6)	47.7(47.3,48.1)	45.2(44.4,45.9)	< 0.001
BMI	29.2(29.1,29.3)	29.2(29.1,29.3)	32.9(32.5,33.4)	< 0.001
Gender				< 0.001
Male	4361(46.0)	3530(45.0)	831(50.7)
Female	5114(54.0)	4307(55.0)	807(49.3)
Race/ethnicity				< 0.001
Mexican American	1836(19.4)	1420(18.1)	416(25.4)
other Hispanic	801(8.5)	642(8.2)	149(9.7)
non-Hispanic white	4246(44.8)	3591(45.8)	655(40.0)
non-Hispanic black	1964(20.7)	1663(21.2)	301(18.4)
other/multiracial	628(6.6)	521(6.7)	107(6.5)
Smoking				0.92
Never	5392(56.9)	4466(57.0)	926(56.5)
Former	2205(23.3)	1816(23.2)	387(23.6)
Current	1880(19.8)	1555(19.8)	325(19.8)
Physical activity				0.03
Heavy	2220(23.4)	1854(23.6)	366(22.3)
Moderate	2564(27.1)	2152(27.5)	412(25.2)
Light	4691(49.5)	3831(48.9)	860(52.5)
Survey cycle				0.2
1999–2000	300(3.2)	245(3.1)	55(3.6)
2001–2002	614(6.5)	513(6.5)	101(6.2)
2003–2004	979(10.3)	800(10.2)	179(10.9)
2005–2006	1762(18.6)	1481(18.9)	281(17.2)
2007–2008	2015(21.3)	1649(21.0)	366(22.3)
2009–2010	2142(22.6)	1750(22.3)	392(23.9)
2011–2012	1663(17.6)	1399(17.9)	264(16.1)
Diabetes				< 0.001
Yes	1490(15.7)	1135(14.5)	355(21.7)
No	7985(84.3)	6702(85.5)	1283(78.3)
Education level				0.1
<High school	2721(28.7)	2219(28.3)	502(30.6)
High school	2191(23.1)	1810(23.1)	381(23.3)
≥college	4563(48.2)	3808(48.6)	755(46.1)
ALT, U/L	25.3(24.9,25.7)	20.3(20.1,20.4)	49.1(47.2,51.0)	< 0.001
AST, U/L	25.3(25.0,25.7)	22.2(22.1,22.3)	40.7(38.9,42.4)	< 0.001
GGT, U/L	27.1(26.4,27.7)	22.3(21.9,22.7)	49.9(47.1,52.7)	< 0.001
TG, U/L	1.75(1.72,1.78)	1.66(1.63,1.69)	2.20(2.10,2.31)	< 0.001
^aAbbreviations: BMI, body mass index. ALT, alanine aminotransferase. AST, aspartate aminotransferase. GGT, gamma glutamyl transferase. TG, triglycerides. NAFLD, nonalcoholic fatty liver disease. NHANES, National Health and Nutrition Examination Survey. ^bAll estimates were accounted for complex survey designs. ^cA total of 107 participants had missing information on BMI.

3.2 Distribution of Blood THMs

Among all 9,475 participants, TCM, BDCM, DBCM, and TBM were measured in 8,948 (94.4%), 9,371 (98.9%), 9,264 (97.8%), and 9,239 (97.5%) participants, respectively. The detection rates were 8,141 (91.0%) for TCM, 7,053 (75.3%) for BDCM, 5,138 (55.5%) for DBCM, and 3,014 (32.6%) for TBM, respectively. Additionally, Br-THM, TTHM, and Cl-THM levels could be calculated for 8,998, 8,522, and 8,683 participants, respectively. The median blood concentrations were 9.2 pg/mL for TCM, 1.5 pg/mL for BDCM, 0.7 pg/mL for DBCM, 0.7 pg/mL for TBM, 3.5 pg/mL for Br-THMs, 14.4 pg/mL for TTHMs, and 13.2 pg/mL for Cl-THMs (Table 2).

Table 2

Distribution of Blood THM Concentrations^a.
	N	N (%) > LOD	Mean	median (IQR)	GM
TCM (pg/mL)	8948	8141 (91.0)	21.92	9.2 (4.4,18.9)	9.27
BDCM (pg/ml)	9371	7053 (75.3)	3.00	1.5 (0.438,3.515)	1.60
DBCM (pg/ml)	9264	5138 (55.5)	2.04	0.70 (0.438,1.9)	0.99
TBM (pg/ml)	9239	3014 (32.6)	2.15	0.71 (0.707,1.20)	1.07
Br-THM (pg/ml)	8998	NA	7.07	3.47 (194,7.04)	4.14
TTHM (pg/ml)	8522	NA	29.01	14.4 (7.64,27.03)	14.93
Cl-THM (pg/ml)	8683	NA	27.14	13.20(6.6,25.2)	13.24
^aAbbreviations: TCM, chloroform. BDCM, bromodichloromethane. DBCM, dibromochloromethane. TBM, bromoform. Br-THMs, the sum of BDCM, DBCM, and TBM. Cl-THMs, the sum of TCM, BDCM, and DBCM. TTHMs, the sum of TCM and Br-THMs. LOD, limit of detection. IQR, interquartile range. GM, geometrical mean. NA, not applicable.

3.3 Blood THMs and NAFLD

To investigate the relationship between blood THM concentration and NAFLD, we constructed a multiple logistic regression model. In model 1, without adjusting for covariates, the concentrations of blood DBCM, TBM, Br-THM, and TTHM were positively associated with the occurrence of NAFLD compared to the control group, with ORs 1.21 (95% CI 1.06–1.37), 1.31 (95% CI 1.12–1.53), 1.22 (95% CI 1.05–1.42), and 1.22 (95% CI 1.04–1.43), respectively. In model 2, after adjusting for age, sex, race, and survey cycles, blood TBM and TTHM concentrations were still significantly positively associated with NAFLD, with ORs 1.20 (95% CI 1.02–1.42) and 1.20 (95% CI 1.01–1.41), respectively. In model 3, after further adjusting for age, sex, race, survey cycles, education level, smoking status, physical activity status, and BMI, the positive correlation of blood TBM and TTHM concentrations with NAFLD remained robust, with ORs 1.27 (95% CI 1.07–1.50) and 1.27 (95% CI 1.07–1.51), respectively. In addition, Br-THM also showed a positive correlation with NAFLD, with an OR 1.19 (95% CI 1.01–1.40). Although the OR values of blood TCM and Cl-THM concentrations contained 1, the p for trend was < 0.05, suggesting that the risk of NAFLD was on the rise with the increase of blood TCM and Cl-THM concentrations. In fully adjusted model 4, blood TBM, Br-THM, and TTHM concentrations were still positively correlated with NAFLD, with ORs 1.27 (95% CI 1.07–1.50), 1.19 (95% CI 1.01–1.40), and 1.27 (95% CI 1.07–1.52), respectively. Furthermore, as the blood concentrations of TCM and Cl-THM increased, the increased trend of NAFLD risk remained evident (Table 3).

Table 3

ORs (95% CIs) of NAFLD to blood THM concentrations based on Multinomial Logistic Regression Model^a.
	OVERALL	Non-NAFLD(%)	NAFLD(%)	Model 1: OR (95% CI)	Model 2: OR (95% CI)	Model 3: OR (95% CI)	Model 4:(95% CI)
TCM (pg/mL)
Quartile 1	2237	1856(83.0)	381(17.0)	Reference	Reference	Reference	Reference
Quartile 2	2237	1869(83.5)	368(16.5)	0.96(0.82,1.12)	0.97(0.82,1.13)	0.98(0.83,1.15)	0.98(0.83,1.15)
Quartile 3	2237	1844(82.4)	393(17.6)	1.04(0.89,1.21)	1.06(0.90,1.24)	1.10(0.93,1.29)	1.10(0.93,1.30)
Quartile 4	2237	1834(82.0)	403(18.0)	1.07(0.92,1.25)	1.09(0.93,1.29)	1.15(0.97,1.37)	1.15(0.97,1.36)
p for trend				0.26	0.16	0.047	0.048
BDCM (pg/mL)
Tertile 1	3124	2601(83.3)	523(16.7)	Reference	Reference	Reference	Reference
Tertile 2	3124	2604(83.4)	520(16.6)	0.99(0.87,1.13)	0.98(0.86,1.12)	1.02(0.88,1.17)	1.02(0.88,1.17)
Tertile 3	3123	2546(81.5)	577(18.5)	1.13(0.99,1.28)	1.09(0.95,1.24)	1.13(0.98,1.30)	1.13(0.99,1.30)
p for trend				0.07	0.21	0.08	0.07
DBCM (pg/mL)
<50%	4632	3893(84.0)	739(16.0)	Reference	Reference	Reference	Reference
50–75%	2318	1888(81.4)	430(18.6)	1.20(1.05,1.37)	1.15(1.00,1.31)	1.15(1.00,1.32)	1.15(1.00,1.32)
>75%	2314	1883(81.4)	431(18.6)	1.21(1.06,1.37)	1.11(0.97,1.27)	1.13(0.98,1.30)	1.13(0.98,1.30)
p for trend				0.002	0.075	0.05	0.05
TBM (pg/mL)
<75%	6932	5772(83.3)	1160(16.7)	Reference	Reference	Reference	Reference
75-87.5%	1156	960(83.0)	196(17.0)	1.02(0.86,1.20)	0.98(0.83,1.17)	0.98(0.82,1.17)	0.98(0.82,1.17)
>87.5%	1151	911(79.1)	240(20.9)	1.31(1.12,1.53)	1.20(1.02,1.42)	1.27(1.07,1.50)	1.27(1.07,1.50)
p for trend				0.002	0.05	0.02	0.01
Br-THM (pg/mL)
Quartile 1	2250	1886(83.8)	364(16.2)	Reference	Reference	Reference	Reference
Quartile 2	2250	1877(83.4)	373(16.6)	1,03(0,88,1.21)	1.01(0.86,1.18)	1.04(0.88,1.23)	1.04(0.88,1.23)
Quartile 3	2249	1860(82.7)	389(17.3)	1.08(0.93,1.27)	1.04(0.89,1.23)	1.05(0.89,1.24)	1.05(0.89,1.25)
Quartile 4	2249	1820(80.9)	429(19.1)	1.22(1.05,1.42)	1.13(0.96,1.32)	1.19(1.01,1.40)	1.19(1.01,1.40)
p for trend				0.008	0.11	0.046	0.043
Cl-THM(pg/mL)
Quartile 1	2171	1813(83.5)	358(16.5)	Reference	Reference	Reference	Reference
Quartile 2	2171	1812(83.5)	359(16.5)	1.00(0.85,1.18)	0.99(0.84,1.16)	1.01(0.85,1.19)	1.01(0.85,1.19)
Quartile 3	2171	1789(82.4)	382(17.6)	1.08(0.92,1.27)	1.08(0.92,1.27)	1.15(0.97,1.36)	1.15(0.97,1.36)
Quartile 4	2170	1773(81.7)	397(18.3)	1.13(0.97,1.33)	1.11(0.95,1.31)	1.17(0.98,1.39)	1.17(0.99,1.39)
p for trend				0.07	0.12	0.03	0.03
TTHM (pg/mL)
Quartile 1	2131	1790(84.0)	341(16.0)	Reference	Reference	Reference	Reference
Quartile 2	2131	1784(83.7)	327(16.3)	1.02(0.87,1.20)	1.01(0.85,1.19)	1.03(0.87,1.22)	1.03(0.86,1.22)
Quartile 3	2130	1749(82.1)	381(17.9)	1.14(0.97,1.34)	1.13(0.96,1.34)	1.19(1.01,1.42)	1.20(1.01,1.42)
Quartile 4	2130	1729(81.8)	401(18.8)	1.22(1.04,1.43)	1.20(1.01,1.41)	1.27(1.07,1.51)	1.27(1.07,1.52)
p for trend				0.006	0.014	0.002	0.002
^aAbbreviations: TCM, chloroform. BDCM, bromodichloromethane. DBCM, dibromochloromethane. TBM, bromoform. Br-THMs, the sum of BDCM, DBCM, and TBM. Cl-THMs, the sum of TCM, BDCM, and DBCM. TTHMs, the sum of TCM and Br-THMs. OR, odds ratio; 95%CI, 95% confidence interval. NAFLD, non-alcoholic fatty liver disease. Model 1: no covariates were adjusted. Model 2: age, gender, ethnicity, and survey cycles were adjusted. Model 3: age, gender, race, survey cycles, education, smoking status, physical activity, and BMI were adjusted. Model 4: age, gender, race, survey cycles, education, smoking status, physical activity, BMI and diabetes were adjusted.

Meanwhile, we also performed stratification and interaction analyses of TTHM to minimize the risk of incidental results due to multiple testing. The results showed that there was no significant interaction between blood TTHM concentration and factors such as age, sex, race, BMI, smoking status, alcohol consumption, diabetes mellitus, or physical activity in the development of NAFLD. However, the association between blood TTHM concentration and NAFLD was slightly stronger among participants who were older than 60 years, females, those with below/normal weight, nonsmokers, nondrinkers, and diabetics (Table 4).

Table 4

Stratification and interaction analysis^a.
	TTHM (pg/mL) OR (95% CI)^b
	Quartile 1	Quartile 2	Quartile 3	Quartile 4	P for trend	P for multiplicative interaction
BMI						0.38
Under/normal weight	Reference	1.35(0.91,2.03)	0.93(0.61,1.43)	1.86(1.25,2.76)	0.01
Overweight	Reference	0.98(0.72,1.34)	1.27(0.94,1.71)	1.00(0.73,1.37)	0.59
Obese	Reference	0.93(0.73,1.17)	1.15(0.91,1.45)	1.18(0.93,1.51)	0.06
Gender						0.14
Male	Reference	1.16(0.91,1.48)	1.29(1.01,1.66)	1.23(0.95,1.58)	0.08
Female	Reference	0.90(0.70,1.16)	1.08(0.85,1.38)	1.30(1.02,1.66)	0.01
Age						0.49
20–40	Reference	0.98(0.74,1.30)	1.26(0.95,1.66)	1.25(0.94,1.65)	0.04
41–60	Reference	0.98(0.74,1.29)	1.10(0.83,1.46)	1.09(0.83,1.46)	0.4
> 60	Reference	1.11(0.76,1.61)	1.09(0.75,1.59)	1.68(1.15,2.44)	0.01
Smoking status						0.12
Never	Reference	1.20(0.95,1.51)	1.23(0.98,1.56)	1.38(1.09,1.75)	0.009
Former	Reference	0.61(0.42,0.88)	1.01(0.76,1.52)	1.02(0.71,1.45)	0.32
Current	Reference	1.19(0.81,1.75)	1.24(0.84,1.83)	1.24(0.83, 1.86)	0.27
Drinking status						0.36
Drinking	Reference	0.95(0.62,1.47)	0.80(0.50,1.27)	0.97(0.63,1.52)	0.78
Non-Drinking	Reference	1.05(0.87,1.26)	1.28(1.06,1.54)	1.33(1.10,1.61)	< 0.001
Diabetes						0.17
Yes	Reference	1.29(0.87,1.91)	1.15(0.76,1.73)	1.80(1.19,2.72)	0.02
No	Reference	0.98(0.80,1.18)	1.19(0.99,1.44)	1.19(0.98,1.44)	0.02
Physical activity						0.90
Heavy	Reference	1.26(0.87,1.83)	1.31(0.90,1.92)	1.46(1.01,2.13)	0.05
Moderate	Reference	0.83(0.59,1.16)	1.14(0.82,1.59)	1.23(0.87,1.73)	0.09
Light	Reference	1.03(0.81,1.31)	1.17(0.92,1.48)	1.21(0.95,1.54)	0.08
Race/ethnicity						0.66
Mexican American	Reference	1.09(0.75,1.59)	1.00(0.68,1.46)	1.40(0.98,2.05)	0.09
Other Hispanic	Reference	1.07(0.55,2.09)	1.58(0.82,3.03)	1.34(0.70,2.57)	0.23
Non-Hispanic White	Reference	0.96(0.74,1.25)	1.24(0.96,1.60)	1.26(0.96,1.65)	0.03
Non-Hispanic Black	Reference	0.76(0.51,1.15)	0.94(0.63,1.40)	0.92(0.61,1.37)	0.96
Other/Multiracial	Reference	2.36(1.16,4.81)	1.70(0.79,3.65)	1.56(0.70,3.45)	0.51
^aAbbreviations: TTHMs, the sum of TCM, BDCM, DBCM, and TBM. OR, odds ratio. 95% CI, 95% confidence interval. BMI, Body Mass Index. ^bAdjusted for age, gender, race, survey cycles, education, smoking status, physical activity, BMI and diabetes.

3.4. Sensitivity analyses

After excluding participants with missing BMI data (n = 107) or who had used swimming pools, hot tubs, or steam rooms in the past 72 h (n = 260), the association between blood THM concentrations and NAFLD remained stable. This consistency persisted after further adjustments for the time interval since the last shower or bath and for co-exposure to other THMs, which were included as covariates in the multivariate analysis (Table 5).

Table 5

Sensitivity analysis ^a,
	OR (95% CI)
	Excluding participants with missing BMI data n = 107^b	Excluding participants who used swimming pools, hot tubs, or steam rooms in the past 72 hours n = 260^b		With additional adjustments for the time interval since the last shower or bath (≤ 2, 3–6, 7–14, or > 14 hours)^c		With additional adjustment for the summed concentrations of all other THMs^d
TCM (pg/mL)
Quartile 1	Reference	Reference	Reference		Reference
Quartile 2	0.98(0.83 ,1.16)	1.00(0.85 ,1.19)	1.00(0.84 ,1.18)		0.99(0.84 ,1.17)
Quartile 3	1.09(0.92,1.29)	1.08(0.91,1.28)	1.10(0.93,1.31)		1.09(0.92,1.28)
Quartile 4	1.16(0.98,1.38)	1.16(0.97,1.37)	1.17(0.98,1.39)		1.13(0.96,1.35)
p for trend	0.04	0.07	0.04		0.09
BDCM (pg/mL)
Tertile 1	Reference	Reference	Reference		Reference
Tertile 2	1.03(0.89,1.19)	1.00(0.87,1.16)	1.02(0.89,1.18)		1.02(0.89,1.18)
Tertile 3	1.15(1.00,1.32)	1.14(0.99,1.31)	1.15(0.99,1.32)		1.14(0.99,1.31)
p for trend	0.05	0.07	0.06		0.07
DBCM (pg/mL)
<50%	Reference	Reference	Reference		Reference
50–75%	1.15(1.00,1.32)	1.15(1.00,1.32)	1.16(1.01,1.33)		1.15(0.99,1.33)
>75%	1.13(0.98,1.30)	1.13(0.98,1.31)	1.14(0.99,1.32)		1.11(0.95,1.29)
p for trend	0.06	0.06	0.05		0.12
TBM (pg/mL)
<75%	Reference	Reference	Reference		Reference
75-87.5%	0.96(0.81,1.15)	0.98(0.82,1.18)	0.97(0.81,1.16)		0.97(0.81,1.16)
>87.5%	1.28(1.08,1.51)	1.29(1.08,1.53)	1.28(1.08,1.51)		1.27(1.07,1.51)
p for trend	0.02	0.01	0.02		0.02
Br-THM (pg/mL)
Quartile 1	Reference	Reference	Reference		Reference
Quartile 2	1.04(0.88,1.23)	1.03(0.87,1.22)	1.05(0.88,1.24)		1.05(0.88,1.25)
Quartile 3	1.08(0.91,1.28)	1.05(0.89,1.25)	1.07(0.90,1.27)		1.06(0.89,1.26)
Quartile 4	1.19(1.01,1.41)	1.19(1.01,1.41)	1.21(1.02,1.44)		1.20(1.01,1.42)
p for trend	0.04	0.04	0.03		0.047
Cl-THM (pg/mL)
Quartile 1	Reference	Reference	Reference		Reference
Quartile 2	1.02(0.86,1.21)	1.02(0.86,1.22)	1.02(0.86,1.21)		1.03(0.87,1.22)
Quartile 3	1.14(0.96,1.36)	1.15(0.97,1.37)	1.16(0.98,1.37)		1.14(0.96,1.36)
Quartile 4	1.19(1.00,1.42)	1.19(0.99,1.41)	1.19(1.00,1.42)		1.19(1.00,1.41)
p for trend	0.02	0.03	0.02		0.03
TTHM (pg/mL)
Quartile 1	Reference	Reference	Reference		NA
Quartile 2	1.03(0.87,1.23)	1.06(0.89,1.26)	1.04(0.88,1.24)		NA
Quartile 3	1.21(1.02,1.43)	1.18(0.99,1.40)	1.20(1.01,1.43)		NA
Quartile 4	1.27(1.07,1.52)	1.30(1.09,1.55)	1.31(1.09,1.56)		NA
p for trend	0.002	0.002	0.001		NA
^aAbbreviations: THMs, trihalomethanes, TCM, chloroform, BDCM, bromodichloromethane, DBCM, dibromochloromethane, TBM, bromoform, Br-THMs, the sum of BDCM, DBCM, and TBM, Cl-THMs, the sum of TCM, BDCM, and DBCM, TTHMs, the sum of TCM and Br-THMs, OR, odds ratio, 95%CI, 95% confidence interval, NA, not applicable. ^bAdjusted for age, gender, race, survey cycles, education, smoking status, physical activity, BMI and diabetes. ^cAdjusted for age, gender, race, survey cycles, education, smoking status, physical activity, BMI, diabetes and the time interval since the last shower or bath (≤ 2, 3–6, 7–14, or > 14 h). ^dAdjusted for age, gender, race, survey cycles, education, smoking status, physical activity, BMI, diabetes and the summed concentrations of all other THMs.

This cross-sectional study of a representative 9,475 adults over the age of 20 in USA showed that blood concentrations of TBM, Br-THM, TTHM, and NAFLD were significantly positively associated with NAFLD, and that elevated levels of TCM and Cl-THM were also associated with an increased risk of NAFLD. Additionally, the positive association between TTHM and NAFLD was more pronounced in women, over the age of 60, below/normal weight, non-smokers, non-drinkers, and diabetics.

The previous experimental animal studies have shown that several disinfection by-products, such as dibromoacetic acid (DBAA) in haloacetic acid (HAA), significantly increase the incidence of hepatocellular tumors [37]. Furthermore, rats exposed to dichloroacetic acid (DCAA) and trichloroacetic acid (TCAA) in drinking water exhibited significant hepatotoxicity and liver damage [38, 39]. Seth et al. found that long-term exposure to BDCM could induce non-alcoholic NASH through free radical metabolism, which is co-mediated by the cytochrome P450 isoenzyme CYP2E1 and the adipokine leptin [21]. Dong et al. reported that exposure to THMs in mice led to an increase in hepatic macrophages, resulting in acute liver injury [40]. Epidemiological studies have also confirmed the association between disinfection by-products and liver injury. Burch et al. found that individuals with blood DBCM concentrations above the median had increased serum alanine aminotransferase (ALT) activity, and this association was more pronounced in non-drinking individuals [23]. Yang et al.'s cross-sectional study showed that Br-THMs and TTHMs were significantly and dose-dependently positively correlated with liver function indexes, such as alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin (TBIL), and indirect bilirubin (IBIL) [41]. These findings confirm our results, and we further evaluated the association between blood THM exposure and NAFLD, which is the first systematic study of this association. We also found that blood THM concentrations had a negative impact on NAFLD. Blood TBM, Br-THM, and TTHM concentrations were significantly positively correlated with NAFLD in both unadjusted and fully adjusted models, and remained robust after sensitivity analyses. Blood TCM and Cl-THM also showed a tendency to be associated with NAFLD.

The mechanism of THMs promoting NAFLD is still unclear. THMs may contribute to the occurrence and progression of NAFLD through various mechanisms, such as interfering hepatocyte metabolism, inducing oxidative stress and inflammatory responses, and participating in the regulation of liver fibrosis. Firstly, THMs may induce NAFLD by interfering with glucolipid metabolism in hepatocytes. They can inhibit key glucose-metabolizing enzyme, such as hexokinase and pyruvate kinase, leading to glycolysis disorder, insulin resistance, and promoting lipogenesis, and fatty acid accumulation in the liver [42, 43]. Insulin resistance is a critical factor in the development of NAFLD, triggering lipotoxicity, oxidative stress, and inflammatory cascades [44, 45]. Additionally, studies have shown that THMs can significantly increase oxidative stress in hepatocytes, resulting in increased levels of reactive oxygen species (ROS), and activating protein-1 (AP-1) and nuclear factor-κB (NF-κB) pathways, promoting the release of inflammatory factors from hepatocytes and exacerbating hepatic inflammatory responses [46–48]. Therefore, THMs directly damages hepatocyte function by inducing oxidative stress and inflammation in hepatocytes, further aggravating the development of NAFLD. Furthermore, THMs may indirectly affect the activation of hepatic stellate cells by interfering with hepatocyte apoptosis and the release of inflammatory factors, resulting in a large number of collagen fibers and extracellular matrix synthesis, and accelerating hepatic fibrosis[49].

The present study included a large and nationally representative sample size, measured internal exposure to disinfection by-product (DBP) biomarkers, and used blood THM concentrations for assessmen to avoid the spatiotemporal variability of high THM concentrations in water [17] and included multiple models to systematically and comprehensively assess the association with NAFLD. Additionally, our study spanned 14 years and included seven cycles, filling in a gap regarding the relationship between disinfection by-products and NAFLD. However, there are some limitations in present study. Firstly, the cross-sectional study design only allowed us to observe data at a single point in time, and causality cannot be determined. In this case, we could only observe an association between blood THM exposure and NAFLD, but could not determine whether exposure was a cause or consequence of NAFLD. Secondly, although we have adjusted for numerous potential confounders, undetected confounders, random errors and so on may affect our results due to the complexity of NAFLD development. Thirdly, although blood concentrations were used as more reliable biomarkers of trihalomethane exposure, these concentrations can change over time, so exposure misclassification was still possible. Fourthly, due to the relatively low detection rates of DBCM and TBM concentrations in blood (55.5% and 32.6%, respectively), this may result in biased in risk estimates. Finally, our study did not explore the association of other DBPs with NAFLD, which may limit overall impact of disinfection by-products on NAFLD risk.

The results of this cross-sectional study, based on a nationally representative population, suggested that higher blood THM concentrations were associated with NAFLD in USA adults. This finding indicated that reducing exposure to THMs through improving water quality management could provide new strategy for NAFLD prevention. However, more prospective studies are needed to confirm this conclusion and elucidate the underlying mechanisms in the future.

Data Availability

The data used in the present research were obtained from publicly accessible sources. These data could be accessible at the following URL: https://www.cdc.gov/nchs/nhanes/.

Funding

This work was supported by the Natural Science Foundation of Zhejiang Province (LY24H020008 and LTGY23H030003), the Key Research and Development Program of Zhejiang Province (2023C03018), and the Fourth Batch of Wenzhou Medical University “Outstanding and Excellent Youth Training Project” (604090352/640).

Ethics approval and consent to participate

Ethical review and approval were waived for this study as it solely used publicly available data for research and publication. The NCHS Institutional Review Board has approved NHANES’s investigation, and all participants have provided written informed consent.

Declaration of Competing Interest

The authors have declared that no competing interests exist.

Author Contribution

SZ conceived the project, designed the research and drafted the manuscript. WJ and XG directed this research. SQ, XS, BY and YZ revised the manuscript. All authors contributed to the article and approved the submitted version.

Vernon, G., A. Baranova, and Z.M. Younossi, Systematic review: the epidemiology and natural history of non-alcoholic fatty liver disease and non-alcoholic steatohepatitis in adults. Aliment Pharmacol Ther, 2011. 34(3): p. 274-85.
Friedman, S.L., et al., Mechanisms of NAFLD development and therapeutic strategies. Nat Med, 2018. 24(7): p. 908-922.
Tacke, F., et al., An integrated view of anti-inflammatory and antifibrotic targets for the treatment of NASH. J Hepatol, 2023. 79(2): p. 552-566.
Huang, D.Q., H.B. El-Serag, and R. Loomba, Global epidemiology of NAFLD-related HCC: trends, predictions, risk factors and prevention. Nat Rev Gastroenterol Hepatol, 2021. 18(4): p. 223-238.
Yang, W., et al., PPARα/ACOX1 as a novel target for hepatic lipid metabolism disorders induced by per- and polyfluoroalkyl substances: An integrated approach. Environ Int, 2023. 178: p. 108138.
Younossi, Z.M., et al., Global epidemiology of nonalcoholic fatty liver disease-Meta-analytic assessment of prevalence, incidence, and outcomes. Hepatology, 2016. 64(1): p. 73-84.
Ruhl, C.E. and J.E. Everhart, Fatty liver indices in the multiethnic United States National Health and Nutrition Examination Survey. Aliment Pharmacol Ther, 2015. 41(1): p. 65-76.
Pierantonelli, I. and G. Svegliati-Baroni, Nonalcoholic Fatty Liver Disease: Basic Pathogenetic Mechanisms in the Progression From NAFLD to NASH. Transplantation, 2019. 103(1): p. e1-e13.
Guo, X., et al., Non-Alcoholic Fatty Liver Disease (NAFLD) Pathogenesis and Natural Products for Prevention and Treatment. Int J Mol Sci, 2022. 23(24).
Tanase, D.M., et al., The Intricate Relationship between Type 2 Diabetes Mellitus (T2DM), Insulin Resistance (IR), and Nonalcoholic Fatty Liver Disease (NAFLD). J Diabetes Res, 2020. 2020: p. 3920196.
Muzurović, E., D.P. Mikhailidis, and C. Mantzoros, Non-alcoholic fatty liver disease, insulin resistance, metabolic syndrome and their association with vascular risk. Metabolism, 2021. 119: p. 154770.
Bambha, K., et al., Ethnicity and nonalcoholic fatty liver disease. Hepatology, 2012. 55(3): p. 769-80.
Zheng, S., et al., Effects of environmental contaminants in water resources on nonalcoholic fatty liver disease. Environ Int, 2021. 154: p. 106555.
Dolce, A. and S. Della Torre, Sex, Nutrition, and NAFLD: Relevance of Environmental Pollution. Nutrients, 2023. 15(10).
Meroni, M., et al., Nutrition and Genetics in NAFLD: The Perfect Binomium. Int J Mol Sci, 2020. 21(8).
Sun, Y., et al., Association of blood trihalomethane concentrations with asthma in US adolescents: nationally representative cross-sectional study. Eur Respir J, 2022. 59(5).
Rivera-Núñez, Z., et al., Comparison of trihalomethanes in tap water and blood: a case study in the United States. Environ Health Perspect, 2012. 120(5): p. 661-7.
Komulainen, H., Experimental cancer studies of chlorinated by-products. Toxicology, 2004. 198(1-3): p. 239-48.
Tafesse, N., et al., Exposure and carcinogenic risk assessment of trihalomethanes (THMs) for water supply consumers in Addis Ababa, Ethiopia. Toxicol Rep, 2023. 10: p. 261-268.
Das, S., et al., Proinflammatory adipokine leptin mediates disinfection byproduct bromodichloromethane-induced early steatohepatitic injury in obesity. Toxicol Appl Pharmacol, 2013. 269(3): p. 297-306.
Seth, R.K., et al., Environmental toxin-linked nonalcoholic steatohepatitis and hepatic metabolic reprogramming in obese mice. Toxicol Sci, 2013. 134(2): p. 291-303.
Makris, K.C., et al., Association between exposures to brominated trihalomethanes, hepatic injury and type II diabetes mellitus. Environ Int, 2016. 92-93: p. 486-93.
Burch, J.B., et al., Trihalomethane exposure and biomonitoring for the liver injury indicator, alanine aminotransferase, in the United States population (NHANES 1999-2006). Sci Total Environ, 2015. 521-522: p. 226-34.
NCHS About the National Health and Nutrition Examination Survey. . June 15, 2024 [cited June 15, 2024; Available from: https://www.cdc.gov/nchs/nhanes/about_nhanes.htm.
NCHS National Health and Nutrition Examination Survey. Nov 30, 2021; Available from: https://www.cdc.gov/nchs/nhanes/index.htm?CDC_AA_refVal=.
NHANES 2011-2012 Procedure Manuals. 2012; Available from: https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/manuals.aspx?BeginYear=2012.
NCHS Laboratory Procedure Manual. Trihalomethanes/MTBE/Nitromethane. April 30, 2021; Available from: https://wwwn.cdc.gov/nchs/data/nhanes/2011-2012/labmethods/VOCMWB_G_MET.pdf.
NCHS 2011-2012 Data Documentation, Codebook, and Frequencies. Volatile Organic Compounds (VOCs) - Trihalomethanes/MTBE/Nitromethane - Blood (VOCMWB_G). 2024; Available from: https://wwwn.cdc.gov/Nchs/Nhanes/2011-2012/VOCMWB_G.htm.
Li, L., et al., The Association between Non-Alcoholic Fatty Liver Disease (NAFLD) and Advanced Fibrosis with Serological Vitamin B12 Markers: Results from the NHANES 1999-2004. Nutrients, 2022. 14(6).
Kallwitz, E.R., et al., Prevalence of suspected nonalcoholic fatty liver disease in Hispanic/Latino individuals differs by heritage. Clin Gastroenterol Hepatol, 2015. 13(3): p. 569-76.
Ong, J.P., A. Pitts, and Z.M. Younossi, Increased overall mortality and liver-related mortality in non-alcoholic fatty liver disease. J Hepatol, 2008. 49(4): p. 608-12.
Chalasani, N., et al., The diagnosis and management of non-alcoholic fatty liver disease: practice guideline by the American Gastroenterological Association, American Association for the Study of Liver Diseases, and American College of Gastroenterology. Gastroenterology, 2012. 142(7): p. 1592-609.
Pan, J., et al., Association between Dietary Niacin Intake and Nonalcoholic Fatty Liver Disease: NHANES 2003-2018. Nutrients, 2023. 15(19).
Carson, A.P., et al., Comparison of A1C and fasting glucose criteria to diagnose diabetes among U.S. adults. Diabetes Care, 2010. 33(1): p. 95-7.
Sun, Y., et al., Blood Trihalomethane Concentrations and Osteoarthritis among U.S. Population Aged over 50 Years. Environ Sci Technol, 2023. 57(43): p. 16166-16175.
Tan, L., et al., Association of iron status with non-alcoholic fatty liver disease and liver fibrosis in US adults: a cross-sectional study from NHANES 2017-2018. Food Funct, 2023. 14(12): p. 5653-5662.
Melnick, R.L., et al., Toxicity and carcinogenicity of the water disinfection byproduct, dibromoacetic acid, in rats and mice. Toxicology, 2007. 230(2-3): p. 126-36.
Gong, T., et al., Dibromoacetic Acid Induced Hepatotoxicity in Mice through Oxidative Stress and Toll-Like Receptor 4 Signaling Pathway Activation. Oxid Med Cell Longev, 2019. 2019: p. 5637235.
Arem, A.E., et al., Date palm fruit extract attenuated oxidative stress induced by two haloacetic acids in Wistar rats. Mediterranean journal of nutrition and metabolism, 2017. 10(2): p. 141-152.
Dong, F., et al., Evidence-based analysis on the toxicity of disinfection byproducts in vivo and in vitro for disinfection selection. Water Res, 2019. 165: p. 114976.
Yang, L., et al., Urinary trihalomethane concentrations and liver function indicators: a cross-sectional study in China. Environ Sci Pollut Res Int, 2023. 30(14): p. 39724-39732.
Bugianesi, E., et al., Insulin resistance in nonalcoholic fatty liver disease. Curr Pharm Des, 2010. 16(17): p. 1941-51.
Guilherme, A., et al., Adipocyte dysfunctions linking obesity to insulin resistance and type 2 diabetes. Nat Rev Mol Cell Biol, 2008. 9(5): p. 367-77.
Peverill, W., L.W. Powell, and R. Skoien, Evolving concepts in the pathogenesis of NASH: beyond steatosis and inflammation. Int J Mol Sci, 2014. 15(5): p. 8591-638.
Buzzetti, E., M. Pinzani, and E.A. Tsochatzis, The multiple-hit pathogenesis of non-alcoholic fatty liver disease (NAFLD). Metabolism, 2016. 65(8): p. 1038-48.
Abbassi, R., N. Chamkhia, and M. Sakly, Chloroform-induced oxidative stress in rat liver: Implication of metallothionein. Toxicol Ind Health, 2010. 26(8): p. 487-96.
Hassoun, E.A., J. Cearfoss, and J. Spildener, Dichloroacetate- and trichloroacetate-induced oxidative stress in the hepatic tissues of mice after long-term exposure. J Appl Toxicol, 2010. 30(5): p. 450-6.
Beddowes, E.J., S.P. Faux, and J.K. Chipman, Chloroform, carbon tetrachloride and glutathione depletion induce secondary genotoxicity in liver cells via oxidative stress. Toxicology, 2003. 187(2-3): p. 101-15.
Kamada, Y., T. Takehara, and N. Hayashi, Adipocytokines and liver disease. J Gastroenterol, 2008. 43(11): p. 811-22.

No competing interests reported.

Download PDF

Reviews received at journal
13 Nov, 2024
Reviewers agreed at journal
23 Oct, 2024
Reviewers invited by journal
06 Oct, 2024
Editor invited by journal
06 Sep, 2024
Editor assigned by journal
05 Sep, 2024
Submission checks completed at journal
05 Sep, 2024
First submitted to journal
04 Sep, 2024

You are reading this latest preprint version

Positive Association Between Blood Trihalomethane Concentrations and Non-Alcoholic Fatty Liver Disease: A Cross-Sectional Study (NHANES)

Status:

Version 1

Abstract

Figures

1. INTRODUCTION

2. METHODS

2.1. Study population

2.2. Blood THM measurements

2.3 Definition of NAFLD

2.4. Assessment of Covariates

2.5. Statistical Analysis

3. RESULTS

3.1. Population Characteristics.

3.2 Distribution of Blood THMs

3.3 Blood THMs and NAFLD

3.4. Sensitivity analyses

4. DISCUSSION

Declarations

Author Contribution

References

Additional Declarations

Status:

Version 1