Gut microbiota involvement in the effect of water-soluble dietary fiber on fatty liver and fibrosis

Abstract

The beneficial effects of water-soluble dietary fiber on liver fat and fibrosis involve the gut microbiota; however, few epidemiological studies have investigated this association. This large-scale epidemiological study aimed to determine the effect of water-soluble dietary fiber intake on liver fat and fibrosis via gut microbiota for the general population. We divided low- and high-intake groups by median daily water-soluble dietary fiber intake and matched background factors by propensity score matching for sex and age. The high-intake group had lower controlled attenuation parameters, a lower fatty liver index, and a lower mac-2-binding protein glycosylated isomer level than the low-intake group. Furthermore, in the high-intake group, the prevalences of metabolic dysfunction-associated steatotic liver disease and cardiometabolic criteria were significantly lower than in the low-intake group. In the high-intake group, there were increases and decreases in 16 bacterial species. Of them, those belonging to Faecalibacterium and Gemmiger had higher relative abundances than the other species and had a negative correlation with the fatty liver index and its components triglyceride and glutamyl transpeptidase in a multivariate analysis adjusted for confounding factors. On the other hand, Dorea showed a significant negative correlation with liver stiffness measure, even though Dorea was decreased in the high-intake group. Faecalibacterium and Gemmiger are butyrate-producing bacteria, suggesting that water-soluble dietary fiber may inhibit fatty liver via gut butyric acid production.

INTRODUCTION

Metabolic dysfunction-associated steatotic liver disease (MASLD) is considered a hepatic phenotype of lifestyle-related diseases with a prevalence worldwide of approximately 30% that is increasing [1]. Diagnosis of MASLD is based on the presence of fatty liver plus one or more of the five cardiometabolic criteria and the exclusion of other fatty liver causes [2]. Fatty liver can lead to fibrosis and is a risk factor for liver cancer, varicose veins, and sarcopenia. Liver fat and fibrosis are closely related to the daily diet, and the Mediterranean diet is known to be effective in preventing the development and progression of MASLD [3, 4]. Furthermore, the gut microbiota is involved in the development of liver fat and fibrosis, and the association between the gut microbiota and the liver is termed the gut-liver axis [5, 6].

Dietary fiber is classified as water-soluble or insoluble based on its solubility in the gastrointestinal tract. Water-soluble dietary fiber is fermented by gut microbiota and converted into short-chain fatty acids such as butyric acid [7, 8]. Dietary fiber is known to effectively prevent lifestyle-related diseases by improving obesity, lipid metabolism, glucose tolerance, and arteriosclerosis [9,10,11]. In addition, several epidemiological studies have reported that dietary fiber might be beneficial in reducing MASLD onset and progression and have also implicated the gut microbiota as a possible mechanism of this beneficial effect [12,13,14,15].

Although many studies have investigated the association between dietary fiber and liver fat and fibrosis, few have investigated the effects of water-soluble dietary fiber, including those of gut microbiota, on liver fat and fibrosis. Furthermore, because the gut microbiota is greatly influenced by confounding factors, such as sex and age, epidemiological studies have yielded different results. This study aimed to epidemiologically investigate the effect of water-soluble dietary fiber intake on MASLD via the gut microbiota in the general population.

MATERIALS AND METHODS

Study participants

This study was conducted as a part of the Iwaki Health Promotion Project. The Iwaki Health Promotion Project is a community-based health promotion project for Japanese residents designed to prevent the onset and progression of lifestyle-related diseases and prolong lifespans. This project is conducted every June as a medical check-up for residents of the Iwaki region of Hirosaki City in Aomori Prefecture, which is in northern Japan [16]. All participants voluntarily participated in response to public announcements. A total of 1,059 adults aged 19–88 years participated in this project, which was conducted in June 2018 (Fig. 1). Of them, 387 participants who were positive in hepatitis B surface antigen (HBsAg) or hepatitis C virus antibody (anti-HCV) testing; had excessive alcohol intakes (male, >30 g/day; female, >20 g/day); had a failure of transient elastography measurement; were taking gastric acid suppressants, steroids, or methotrexate; had a history of gastric surgery; or had missing data were excluded. The remaining participants were divided into low- (336 participants) and high-intake (336 participants) groups based on their median water-soluble dietary fiber intakes per 1,000 kcal (1.51 g/day). Because there were significant differences in sex and age between the two groups, propensity score matching was performed for sex and age to equalize the background factors of both groups. After propensity score matching, 518 participants (259 each in the low- and high-intake groups) were selected, and 511 participants were ultimately included in the analysis.

Fig. 1.

Study enrollment flowchart.

HBsAg: hepatitis B surface antigen; anti-HCV test: hepatitis C virus antibody.

Transient elastography

Transient elastography with liver stiffness measure (LSM) and controlled attenuation parameter (CAP) measurements were performed using a FibroScan Compact 530 device (Echosens, Paris, France) equipped with M and KL probes. Five well-trained hepatology specialists performed all examinations. When the number of measurements was less than 10 or the ratio of the interquartile range was greater than 0.30, the measure values were considered unreliable and excluded.

Clinical parameters

Water-soluble dietary fiber intake was calculated based on the results of the Brief Self-Administered Diet History Questionnaire (BDHQ), a convenient diet assessment questionnaire developed in Japan. The BDHQ is a 4-page self-administered questionnaire that asks about the consumption frequency of selected foods to estimate the dietary intake of 58 commonly consumed food and beverage items in Japan over one month [17]. Participants were given the BDHQ in advance, and each was interviewed individually on the day of the project. Questionnaires were collected after the participants’ answers were confirmed.

The following clinical parameters were recorded on the same day as the transient examination: sex; age; height; body mass index (BMI, calculated by dividing the weight in kilograms by the squared height in meters); waist circumference; HBsAg or anti-HCV test results; and levels of aspartate aminotransferase, alanine aminotransferase, gamma-glutamyl transpeptidase (GGT), glucose, hemoglobin A1c (HbA1c), high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, triglycerides, and Mac-2-binding protein glycosylated isomer (M2BPGi).

The fatty liver index was calculated as follows:

{e(0.953 × ln(triglycerides) + 0.139 × BMI + 0.718 × ln(GGT) + 0.053 × waist circumference − 15.745)}/{1 + e (0.953 × ln(triglycerides) + 0.139 × BMI + 0.718 × ln(GGT) + 0.053 × waist circumference − 15.745)} × 100.

Based on a previous report, participants who had a fatty liver according to a Cap ≥248 dB/m and met any of the following criteria were diagnosed with MASLD: BMI ≥23 kg/m² or a high waist circumference (≥ 90 cm for men and ≥85 cm for women; fasting serum glucose ≥100 mg/dL or type 2 diabetes or a prescription record of antidiabetic medications; blood pressure ≥130/85 mm Hg or a prescription record of antihypertensive medications; triglycerides ≥150 mg/dL or a prescription record of lipid-lowering medications; low HDL cholesterol (≤40 mg/dL for men and ≤50 mg/dL for women) or a prescription record of lipid-lowering medications [2].

Next-generation sequence analysis of gut microbiota

Fecal samples were collected in dedicated containers and suspended in guanidine thiocyanate solution (100 mM Tris-HCl, pH 9.0; 40 mM Tris-ethylenediaminetetraacetic acid [EDTA], pH 8.0; and 4 M guanidine thiocyanate; TechnoSuruga Laboratory Co., Ltd., Shizuoka, Japan). The samples were kept at −80°C before DNA extraction. A series of representative bacterial species in the human gut microbiota were analyzed using primers for the V3–V4 region of the 16S rDNA of prokaryotes in accordance with previous studies [18]. Sequencing was performed using an Illumina MiSeq system (Illumina, San Diego, CA, USA). The methods for quality filtering of the sequences were as follows: only reads with quality value scores ≥0 for more than 99% of the sequences were extracted for the analysis. Detection and identification of the bacteria from the sequences were performed using Metagenome@KIN software version 2.2.1 (World Fusion Co., Tokyo, Japan) and TechnoSuruga Lab Microbial Identification database DB-BA 10.0 (TechnoSuruga Laboratory Co., Ltd., Shizuoka, Japan) at 97% sequence similarity. Relative abundance is presented as the percentage of reads for each bacterium relative to the total number of reads.

Statistical analysis

Categorical variables are presented as frequencies, and continuous variables are presented as medians, along with interquartile ranges. χ² and Mann–Whitney U tests were used to compare the two groups. Microbiota were compared using linear discriminant analysis effect size (LEfSe) analyses [19]. Spearman’s rank correlation coefficients were used to investigate the correlation between fatty liver or fibrosis and gut microbiota. A multiple regression model including the CAP, fatty liver index (FLI), LSM, M2BPGi, and gut microbiota was used for predictive analysis. The independent variables included age, sex, smoking and exercise habits, fasting serum glucose levels, and systolic blood pressure. We controlled for these confounding factors to clarify the relationship between liver fat or fibrosis and gut microbiota. Before multiple regression analysis of fatty liver or fibrosis and gut microbiota, all continuous parameters were log-transformed (natural logarithm) to approximate a normal distribution. Furthermore, the gut microbiota was log-transformed after adding 1 because a relative abundance of 0% was included.

Statistical analyses were performed using the R software (R Foundation for Statistical Computing, version R-4.1.1) and IBM SPSS Statistics version 28.0 (IBM Corp., Armonk, NY, USA). Statistical significance was set at p<0.05.

Ethics statement

This study was conducted in accordance with the ethical standards of the Declaration of Helsinki and was approved by the Hirosaki University Medical Ethics Committee (authorization number, 2018-012; ethical approval date, May 11, 2018). Informed consent was obtained from all participants.

RESULTS

Participant characteristics

The baseline and propensity score-matched characteristics of the participants are presented in Table 1. Although there were significant differences in sex and age at baseline, the two groups showed no significant differences in sex or age after propensity score matching. The median water-soluble dietary fiber intake per 1,000 kcal was 1.20 g/day in the low-intake group and 1.84 g/day in the high-intake group. The high-intake group had significantly lower BMIs, waist circumferences, CAP levels, and FLI, M2BPGi, systolic blood pressure, GGT, and triglyceride levels and had higher HDL cholesterol levels and percentages of exercise habit than the low-intake group. In addition, the prevalence of MASLD was significantly lower in the high-intake group (15.8%) compared with that (32.4%) in the low-intake group.

Table 1. Baseline and propensity score-matched characteristics of the participants.

Variables	Baseline			After propensity score matching
	Low intake group	High intake group	p-value	Low intake group	High intake group	p-value
	N=336	N=336		N=259	N=259
sex, male	134 (39.9)	93 (27.7)	0.001	88 (34.0)	81 (31.3)	0.287
age (year)	44.5 (36.0–60.0)	57.0 (43.3–66.0)	<0.001	51.0 (39.0–64.0)	51.0 (39.0–62.0)	0.774
Water-soluble dietary fiber intake (g/1,000 kcal)	1.19 (1.02–1.33)	1.85 (1.67–2.14)	<0.001	1.20 (1.02–1.33)	1.84 (1.67–2.12)	<0.001
BMI (kg/m2)	22.4 (20.1–24.9)	22.1 (19.9–24.3)	0.277	22.3 (20.2–24.8)	21.5 (19.5–23.4)	0.001
Waist circumference (cm)	83.2 (76.9–89.8)	82.2 (75.3–88.5)	0.133	83.2 (77.0–89.5)	79.5 (74.7–86.6)	<0.001
CAP (dB/m)	224.0 (176.3–265.0)	217.5 (185.0–261.0)	0.770	225.0 (176.0–264.0)	205.0 (173.0–238.0)	0.001
FLI	13.3 (4.8–34.8)	10.8 (5.1–25.6)	0.119	13.7 (5.2–32.5)	8.5 (4.2–18.2)	<0.001
LSM (kPa)	4.3 (3.5–5.3)	4.2 (3.6–5.4)	0.825	4.2 (3.4–5.2)	4.1 (3.5–4.9)	0.594
M2BPGi (C.O.I)	0.50 (0.34–0.68)	0.51 (0.36–0.69)	0.516	0.53 (0.37–0.73)	0.46 (0.34–0.63)	0.008
Systolic blood pressure (mmHg)	121.0 (110.0–131.0)	121.0 (110.0–133.0)	0.620	122.0 (112.0–134.0)	117.0 (109.0–128.0)	0.001
Diastolic blood pressure (mmHg)	75.0 (69.0–83.0)	76.0 (69.0–83.0)	0.881	76.0 (70.0–85.0)	76.0 (68.0–82.0)	0.379
Fasting serum glucose (mg/dL)	90.5 (85.0–97.0)	91.0 (86.0–100.0)	0.055	91.0 (85.0–97.0)	90.0 (85.0–96.0)	0.197
HbA1c (%)	5.6 (5.4–5.8)	5.7 (5.5–5.9)	0.150	5.6 (5.5–5.9)	5.6 (5.4–5.8)	0.156
Aspartate aminotransferase (IU/L)	20.0 (17.0–24.0)	20.5 (17.0–24.0)	0.185	20.0 (17.0–25.0)	20.0 (17.0–24.0)	0.764
Alanine aminotransferase (IU/L)	17.5 (13.0–24.8)	17.0 (13.0–23.0)	0.988	17.0 (13.0–23.0)	16.0 (12.0–22.0)	0.417
GGT (IU/L)	21.0 (15.0–35.0)	19.0 (15.0–29.0)	0.025	20.0 (14.0–34.0)	18.0 (14.0–28.0)	0.018
Triglycerides (mg/dL)	77.0 (51.0–112.8)	74.0 (56.0–102.8)	0.656	79.0 (51.0–112.0)	69.0 (51.0–95.0)	0.045
HDL cholesterol (mg/dL)	61.5 (51.0–74.0)	67.0 (55.0–78.0)	0.001	62.0 (53.0–75.0)	68.0 (55.0–81.0)	0.010
LDL cholesterol (mg/dL)	116.0 (96.8–136.0)	118.0 (101.0–118.0)	0.292	118.0 (98.0–141.0)	116.0 (98.0–136.0)	0.314
smoking habit (%)	50 (14.9)	35 (10.4)	0.104	35 (13.5)	30 (11.6)	0.298
exercise habit (%)	39 (11.6)	76 (22.6)	<0.001	28 (10.8)	53 (20.5)	0.002
MASLD (%)	106 (31.5)	103 (30.7)	0.868	84 (32.4)	41 (15.8)	<0.001

Data are presented as numbers (%) or median (range). BMI: body mass index; CAP: controlled attenuation parameter; LSM: liver stiffness measure; FLI: fatty liver index; M2BPGi: mac-2-binding protein glycosylated isomer; GGT: gamma-glutamyl transpeptidase; HDL: high density lipoprotein; LDL: low density lipoprotein; MASLD: metabolic dysfunction-associated steatotic liver disease.

Figures 2 and 3 show the differences in the composition and diversity of the gut microbiota after propensity score matching. There were no significant differences in the Chao-1 index and principal coordinate analysis. In contrast, the Shannon index was lower in the high-intake group than in the low-intake group.

Fig. 2.

Composition of oral microbiota in the groups with low and high intake of water-soluble dietary fiber. (a) Phyla and (b) Genera levels.

Fig. 3.

Differences in the diversity of gut microbiota between the groups with low and high intake of water-soluble dietary fiber. (a) Chao-1 index, (b) Shannon index, (c) Principal coordinate analysis. *p<0.05.

Comparison of water-soluble dietary fiber intake and gut microbiota

The LEfSe results for water-soluble dietary fiber intake and gut microbiota are shown in Fig. 4. In the high-intake group, we observed increases and decreases in the abundances of eight bacterial species (Fig. 4a, 4b). Among them, the relative abundances of the genera Faecalibacterium, Gemmiger, Dorea, Clostridium XVIII, and Butyricoccus were high, whereas those of Lactobacillus, Catenibacterium, Bacillus, Aeromonas, and Selenomonas were very low (<0.02%; Fig. 4c).

Fig. 4.

linear discriminant analysis effect size (LEfSe) results of the gut microbiota between the groups with low and high intake of water-soluble dietary fiber. (a) Linear discriminant, (b) Cladogram report, (c) Relative abundance histogram of genera levels. LDA: linear discriminant analysis.

Relationship between liver fat or fibrosis and gut microbiota

Table 2 summarizes the results of the correlation between liver fat or fibrosis and gut microbiota associated with water-soluble dietary fiber. Single correlation analysis revealed that Faecalibacterium and Gemmiger had a negative correlation with the FLI. Dorea had a positive correlation with CAP and FLI and a negative correlation with LSM. Lactobacillus had a positive correlation with CAP and M2BPGi.

Table 2. Correlation between liver fat or fibrosis and gut microbiota.

	CAP		FLI		LSM		M2BPGi
	ρ	p-value	ρ	p-value	ρ	p-value	ρ	p-value
Faecalibacterium	−0.073	0.097	−0.118	0.007	−0.074	0.094	−0.014	0.751
Gemmiger	−0.076	0.084	−0.184	<0.001	0.008	0.852	−0.045	0.312
Clostridium XVIII	−0.034	0.443	−0.075	0.088	0.021	0.629	−0.032	0.474
Catenibacterium	0.031	0.483	0.121	0.121	−0.081	0.066	0.070	0.112
Butyricicoccus	0.074	0.092	0.035	0.424	−0.049	0.263	−0.054	0.218
Bacillus	−0.038	0.384	−0.046	0.292	−0.046	0.292	−0.004	0.932
Lactobacillus	0.087	0.049	0.071	0.109	−0.035	0.419	0.112	0.011
Dorea	0.096	0.029	0.103	0.019	−0.095	0.031	0.064	0.147
Aeromonas	−0.040	0.359	−0.043	0.334	0.074	0.094	0.035	0.431
Selenomonas	0.038	0.385	0.042	0.343	0.011	0.800	0.026	0.554

ρ: Spearman’s rank correlation coefficient; CAP: controlled attenuation parameter; FLI: fatty liver index; LSM: liver stiffness measure; M2BPGi: mac-2-binding protein.

Next, multiple regression analysis was performed, with liver fat or fibrosis as the dependent variable and age, sex, smoking habit, exercise habit, fasting serum glucose level, and systolic blood pressure, in addition to gut microbiota, as the independent variables. The results are presented in Table 3. Multiple linear analyses showed a significant negative correlation between the FLI and Faecalibacterium and Gemmiger. Additionally, LSM showed a significant negative correlation with Dorea and Catenibacterium. A positive correlation was also observed between Aeromonas and M2BPGi.

Table 3. Multiple regression analysis of the relationship between liver fat or fibrosis and gut microbiota.

	CAP			FLI			LSM			M2BPGi
	β	p-value	R2	β	p-value	R2	β	p-value	R2	β	p-value	R2
Faecalibacterium	−0.027	0.515	0.149	−0.073	0.040	0.380	−0.065	0.130	0.078	−0.045	0.273	0.152
Gemmiger	−0.025	0.544	0.149	−0.106	0.003	0.385	0.037	0.393	0.075	0.002	0.961	0.150
Clostridium XVIII	−0.017	0.678	0.148	−0.035	0.320	0.376	−0.021	0.628	0.074	−0.044	0.286	0.152
Catenibacterium	−0.034	0.419	0.149	0.043	0.227	0.376	−0.089	0.039	0.082	0.076	0.067	0.156
Butyricicoccus	0.045	0.276	0.150	−0.010	0.781	0.374	−0.021	0.622	0.074	−0.056	0.174	0.153
Bacillus	−0.074	0.073	0.153	−0.062	0.076	0.378	−0.054	0.206	0.077	−0.038	0.360	0.152
Lactobacillus	0.034	0.405	0.149	0.001	0.983	0.374	−0.023	0.600	0.074	0.060	0.147	0.154
Dorea	0.042	0.315	0.150	0.008	0.815	0.374	−0.089	0.038	0.082	0.033	0.421	0.151
Aeromonas	−0.033	0.415	0.149	−0.004	0.899	0.374	0.052	0.221	0.077	0.084	0.040	0.157
Selenomonas	0.037	0.361	0.150	0.050	0.152	0.377	−0.007	0.865	0.074	0.004	0.928	0.150

This multivariate analysis was adjusted for age, sex, smoking habit, exercise habit, fasting serum glucose, and systolic blood pressure. β: standardized coefficient; R²: coefficient of determination; CAP: controlled attenuation parameter; FLI: fatty liver index; LSM: liver stiffness measure; M2BPGi: mac-2-binding protein.

A multiple regression analysis of the relationships of Faecalibacterium and Gemmiger with triglycerides, GGT, BMI, and waist circumference, which constitute the FLI formula, showed significant negative correlation for triglycerides and GGT with both Faecalibacterium and Gemmiger (Table 4).

Table 4. Multiple regression analysis of the relationships of Faecalibacterium and Gemmiger with triglycerides, GGT, BMI, and waist circumference.

	Triglycerides			GGT			BMI			Waist circumference
	β	p	R2	β	p	R2	β	p	R2	β	p	R2
Faecalibacterium	−0.121	0.002	0.244	−0.085	0.029	0.251	−0.029	0.469	0.215	−0.017	0.655	0.272
Gemmiger	−0.147	<0.001	0.250	−0.093	0.017	0.252	−0.043	0.281	0.216	−0.042	0.275	0.270

This multivariate analysis was adjusted for age, sex, smoking habit, exercise habit, fasting serum glucose, and systolic blood pressure. β: standardized coefficient; R²: coefficient of determination; GGT: gamma-glutamyl transpeptidase; BMI: body mass index.

DISCUSSION

In this study, we revealed that high intake of water-soluble dietary fiber reduced liver fat and fibrosis and cardiometabolic criteria associated with MASLD. Furthermore, we found that high intake of water-soluble dietary fiber increased butyric-producing bacteria, such as Faecalibacterium and Gemmiger, and that Faecalibacterium and Gemmiger reduced the FLI and two of its components, triglycerides and GGT.

In this study, after propensity score matching, the group with high intake of water-soluble dietary fiber had significantly fewer patients with MASLD than the low-intake group. The diagnosis of MASLD requires the presence of a fatty liver and one or more of five cardiometabolic criteria [2]. In this study, the high-intake group not only had lower CAP levels and FLIs, an indicator of liver fat content, but also had lower BMIs, waist circumferences, blood pressures, and triglyceride levels and higher HDL cholesterol levels, which are associated with cardiometabolic criteria, than the low-intake group. Water-soluble dietary fiber acts prophylactically against obesity, diabetes mellitus, dyslipidemia, and cardiovascular diseases by improving lipid metabolism and insulin resistance [9,10,11, 20, 21]. In addition, M2BPGi, a specific indicator of liver fibrosis, was lower in the high-intake group than in the low-intake group. Few studies have examined the association between water-soluble dietary fiber and liver fibrosis; however, previous studies have shown that water-soluble dietary fiber reduces cirrhosis-related mortality [22]. This study suggests that water-soluble dietary fiber intake can reduce fatty liver and cardiometabolic criteria, resulting in a lower MASLD risk.

Among the bacterial species that increased in the group with high intake of water-soluble dietary fiber, those belonging to the genera Faecalibacterium, Gemmiger, and Clostridium XVIII were butyrate-producing bacteria [23,24,25]. We have previously reported an increase in Faecalibacterium and Gemmiger in subjects with high intake of water-soluble dietary fiber [26, 27]. Although no significant differences were observed, there was a trend toward lower relative abundances of Faecalibacterium, Gemmiger, and Clostridium XVIII in the MASLD group than in the non-MASLD group (data not shown). Butyrate-producing bacteria use water-soluble dietary fibers as a substrate to produce butyric acid, a short-chain fatty acid. Butyric acid benefits the body via regulatory T cells, preventing obesity, improving insulin resistance and lipid metabolism, and preventing fatty liver and liver fibrosis [28,29,30,31]. Therefore, butyrate-producing bacteria may be an essential mechanism by which water-soluble dietary fibers suppress fatty liver and fibrosis.

In this study, Faecalibacterium and Gemmiger, which increased in the group with high intake of water-soluble dietary fiber, showed a significant negative correlation with FLI, an indicator of liver fat content. A cross-sectional study reported that people with a high FLI have low dietary fiber intake [32, 33]. It has been reported that Faecalibacterium abundance decreases in MASLD [34,35,36]. Butyric acid produced by butyrate-producing bacteria reduces intestinal permeability and inhibits the influx of toxic substances, including lipopolysaccharide, into the liver [30, 31, 37]. Furthermore, butyric acid suppresses MASLD by inhibiting insulin-mediated fat accumulation via regulatory T cells and increasing anti-inflammatory effects [38, 39]. It has also been reported that water-soluble dietary fiber inhibits MASLD progression by decreasing serum zonulin, a tight junction relaxant toxin, thereby improving intestinal permeability; short-chain fatty acids, including butyric acid, have been implicated in this mechanism [40].

On the other hand, no previous studies have shown that Gemmiger is effective for liver fat accumulation and fibrosis. Both Gemmiger and Faecalibacterium belong to the bacterial family Ruminococcaceae [41]. It has been reported that foods rich in water-soluble dietary fibers increase Gemmiger and Faecalibacterium [27]. Gemmiger might effectively prevent liver fat accumulation and fibrosis by increasing butyric acid by the same mechanism as Faecalibacterium.

Multivariate analysis revealed that Faecalibacterium and Gemmiger were significantly and negatively correlated with triglyceride and GGT levels. Butyric acid produced by butyrate-producing bacteria is reported to promote adipogenesis by activating the peroxisome proliferator-activated receptor pathway, thereby lowering triglyceride [42,43,44,45,46].

Dietary fibers, especially fruits, can improve liver-related enzymes [33]. In addition, a randomized controlled trial reported that water-soluble dietary fibers reduced liver-related enzymes by improving intestinal permeability [40]. GGT, a hepatic glycosyltransferase, is closely associated with atherosclerosis and glucose tolerance and is an independent predictor of diabetes mellitus, hypertension, metabolic syndrome, and cardiovascular diseases [47,48,49,50]. GGT is not only involved in the hydrolysis of cellular glutathione but is also a marker of oxidative stress and subclinical inflammation [51]. While increased visceral fat is the cause of elevated liver enzyme levels, GGT itself can also cause chronic liver damage and fatty liver [52, 53]. Decreased Faecalibacterium and elevated GGT levels have been observed in patients with metabolic syndrome [54]. Butyric acid produced by Faecalibacterium is a representative anti-inflammatory molecule and reduces GGT by suppressing the nuclear factor-κB [55,56,57,58]. Similarly, Gemmiger may also reduce GGT through its anti-inflammatory effects via butyric acid production. Although butyric acid was not measured in this study, our results suggest that butyrate-producing bacteria play an essential role in the mechanism by which water-soluble dietary fiber reduces fatty liver.

On the other hand, no significant correlation of BMI or waist circumference, which are both included in the FLI formula, was observed with either Faecalibacterium or Gemmiger. Dietary fiber and butyrate-producing bacteria such as Faecalibacterium, which increase with water-soluble dietary fiber intake, have been reported to prevent obesity by improving anti-inflammatory effects, lipid metabolism disorders, and insulin resistance [59]. We adjusted for confounding factors such as sex, age, and lifestyle; however, BMI and waist circumference showed large individual differences depending on skeletal structure and muscle; as a result, no significant correlation was observed. Furthermore, it has also been reported that a high-quality and healthy diet rich in dietary fiber improves hepatic lipidosis and metabolic dysfunction in patients with MASLD, independent of weight loss [60]. This study suggests that water-soluble dietary fiber may reduce the risk of MASLD by suppressing lipid metabolism and liver inflammation independent of measures of body size, such as BMI and waist circumference.

In this study, Dorea decreased in the group with high intake of water-soluble dietary fibers. Dorea showed no significant association with CAP levels and FLI but a significant negative correlation with LSM. The relative abundance of Dorea in this study participants was relatively high at 0.69%; however, little is known about its effects on humans. Studies in mice have indicated a link between Dorea and chronic stress [61]. A study in humans reported an increase in Dorea following low-sugar diet therapy [62]. Few studies have examined the association between water-soluble dietary fiber and liver fibrosis; however, a previous study has reported that water-soluble dietary fiber intake reduces liver cirrhosis-related mortality [22]. In this study, after propensity score matching, the high-intake group had low M2BPGi, a specific indicator of liver fibrosis. This study showed conflicting results, with Dorea negatively correlated with a liver fibrosis indicator even though it (Dorea) decreased in the high-intake group of water-soluble dietary fiber. The association between Dorea, water-soluble dietary fiber, and liver fibrosis remains unclear and is a topic for future investigation.

This study has several limitations. First, the dietary fiber intake of the participants was only approximately half of the recommended intake. Even the high-intake group did not have adequate intake of water-soluble dietary fiber. Second, the participants were middle-aged and older adults living in rural regions. Therefore, it is inappropriate to generalize our results to younger people or urban residents. Third, butyric acid levels were not measured. Although this study revealed the importance of butyric-producing bacteria, we could only speculate about the disposition of butyric acid.

In conclusion, water-soluble dietary fiber intake can reduce liver fat and fibrosis and MASLD-related cardiometabolic criteria by increasing butyrate-producing bacteria. Water-soluble dietary fiber is an effective nutrient for the prevention of many lifestyle-related diseases.

CONFLICTS OF INTEREST

The authors declare they have no conflicts of interest associated with this manuscript.