Vitamin B6 efficacy in the treatment of nonalcoholic fatty liver disease: an open-label, single-arm, single-center trial

Abstract

Vitamin B6 is an important cofactor in fat metabolism and its deficiency has been correlated with nonalcoholic fatty liver disease. However, no study has investigated the efficacy of vitamin B6 supplementation in these patients. The aim of this open-label, single-arm, single-center study was to examine the therapeutic effect of vitamin B6 in patients with nonalcoholic fatty liver disease. Twenty-two patients with nonalcoholic fatty liver disease received vitamin B6 (90 mg/day) orally for 12 weeks. Clinical parameters were evaluated, and liver fat and fibrosis were quantified before and after treatment using magnetic resonance imaging-based proton density fat fraction and magnetic resonance elastography. Serum alanine aminotransferase levels, the primary endpoint, did not change significantly after vitamin B6 treatment (93.6 ± 46.9 to 93.9 ± 46.6, p = 0.976). On the other hand, magnetic resonance imaging-based proton density fat fraction, a parameter of hepatic lipid accumulation, was significantly reduced (18.7 ± 6.1 to 16.4 ± 6.4, p<0.001) despite no significant changes in body mass index, even in those not taking vitamin E (n = 17, 18.8 ± 6.9 to 16.7 ± 7.3, p = 0.0012). Vitamin B6 administration significantly ameliorated hepatic fat accumulation. As an inexpensive agent with few side effects, vitamin B6 could be a novel therapeutic agent for the treatment of nonalcoholic fatty liver disease.

Introduction

Nonalcoholic fatty liver disease (NAFLD) has become the most common cause of chronic liver disease worldwide with an estimated global prevalence of 25%.⁽¹⁾ NAFLD represents a broad spectrum of disorders, ranging from nonalcoholic fatty liver to nonalcoholic steatohepatitis, cirrhosis, and hepatocellular carcinoma.⁽²⁾ NAFLD is associated with lifestyle-related diseases such as obesity, diabetes, and dyslipidemia and is considered a hepatic component of metabolic syndrome. Although the prevalence of NAFLD has increased with recent lifestyle changes, only a few effective treatments exist.

Vitamin B6 (VitB6) is a generic name that includes 6 vitamers: pyridoxine (PN), pyridoxal (PL), pyridoxamine (PM), and their respective phosphate esters pyridoxine 5'-phosphate (PNP), pyridoxal 5'-phosphate (PLP), and pyridoxamine 5'-phosphate (PMP).⁽³⁾ VitB6 works as a cofactor for more than 150 enzyme reactions involved in amino acid, glucose, and fat metabolism. PLP is the biologically active form of VitB6 and low levels of plasma PLP are associated with cardiovascular disease, stroke, venous thrombosis, rheumatoid arthritis, inflammatory bowel disease, diabetes, and several cancers.^(4–11) VitB6 intake and hepatic steatosis are negatively correlated and patients with NAFLD have been found to have diets low in VitB6 compared to healthy individuals.^(12,13) In addition, patients with NAFLD have low levels of plasma VitB6, and VitB6 administration has been shown to ameliorate hepatic lipid accumulation in mice.⁽¹⁴⁾ Although these studies reveal that VitB6 deficiency may be associated with the progression of NAFLD, there is no known report that has investigated the efficacy of VitB6 in patients with NAFLD. Therefore, this study aimed to examine the therapeutic effect of VitB6 administration in patients with NAFLD.

Materials and Methods

Study design and patients

The study protocol was conducted in accordance with the guidelines contained within the Declaration of Helsinki and was approved by the ethics committees of Yokohama City University (B180405001). Written informed consent was obtained from all participants before study participation. The trial is registered (date of registration: May 16, 2018) with the University Hospital Medical Information Network (UMIN) Clinical Trials Registry (UMIN000032611).

This was an open-label, single-arm, single-center study, and we screened patients with NAFLD who regularly visited the Yokohama City University Hospital (Yokohama, Japan). NAFLD was diagnosed based on ultrasonography findings. The inclusion criteria were as follows: age between 20 and 85 years, ineffective lifestyle intervention for more than 3 months, magnetic resonance imaging proton density fat fraction (MRI-PDFF) value >5.2%, magnetic resonance elastography (MRE) <6.7 kPa, and alanine aminotransferase (ALT) levels >40 IU/L. These MRI-PDFF and MRE values correspond to steatosis grade ≥1 and fibrosis stage ≤3, respectively.⁽¹⁵⁾ Patients with a history of viral hepatitis (caused by hepatitis B, hepatitis C, or Epstein–Barr virus), autoimmune hepatitis, primary biliary cholangitis, sclerosing cholangitis, hemochromatosis, α1-antitrypsin deficiency, Wilson’s disease, drug-induced hepatitis, cirrhosis, substantial alcohol consumption (>20 g/day for women; >30 g/day for men), or any other severe comorbidity were excluded. All subjects took 30 mg of VitB6 (pyridoxine hydrochloride) tablets after each meal three times a day for 12 weeks.

All patients were required not to change their doses of Vitamin E (VitE), any diabetes medications, any dyslipidemia medications, and any antihypertensive medications from 12 weeks prior to the start of oral VitB6 until the end of the treatment. The primary endpoint was the change in ALT levels after VitB6 administration. Secondary endpoints were changes in MRI-PDFF, MRE, routine liver biochemistries, lipid profiles, and serum VitB6 levels from baseline to 12 weeks after VitB6 administration.

Laboratory and clinical parameters

Venous blood samples were obtained after patients fasted for 8 h. Blood cell count, fasting blood sugar (FBS), hemoglobin A1c, aspartate aminotransferase (AST), ALT, γ-glutamyl transpeptidase, alkaline phosphatase, total bilirubin, total cholesterol, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein cholesterol, triglycerides, high sensitivity C-reactive protein, ferritin, blood urea nitrogen, creatinine, and type IV collagen 7s were measured before and after VitB6 treatment. Levels of three vitamers (PN, PL, and PM) were also assessed. Hemoglobin A1c was expressed in National Glycohemoglobin Standardization Program units. The fibrosis-4 (Fib-4) index was calculated as follows: [age (years) × AST (IU/L)]/[platelet count (10⁹/L) ×√ALT (IU/L)].⁽¹⁶⁾ Insulin resistance was assessed by the homeostasis model assessment of insulin resistance [HOMA-IR; IRI (IU/ml) × FBS (mg/dl)/405]. To assess the health-related quality of life outcomes, we used the Short-Form-8 questionnaire, as reported previously.⁽¹⁷⁾ Baseline clinical parameters, except for MRE and MRI-PDFF, were collected on the day of starting the medication, and post-treatment clinical parameters were collected at 12 ± 2 weeks after the start of treatment.

Assessment of liver stiffness and steatosis

All included patients underwent hepatic MRE examinations by 3.0-T imagers (GE Healthcare, Milwaukee, WI). MRI-PDFF was measured by the modified Dixon method with advanced processing (IDEAL IQ, GE Healthcare). Calculations and interpretations of MRE and MRI-PDFF were performed by one author using previously established methods.⁽¹⁵⁾ The author was blinded to the clinical and laboratory findings. Baseline MRE and MRI-PDFF were performed within 12 weeks prior to initiation of VitB6 treatment, and post-treatment MRE and MRI-PDFF were performed at 12 ± 2 weeks after the start of treatment.

Statistical analyses

Data are expressed as mean ± SD unless indicated otherwise. The sample size was determined based on previous reports.⁽¹⁸⁾ Statistical power was set at 90%, with a two-sided type 1 error of 0.05, a change in ALT of –38 and a SD of 47. The number of cases was set at 23 considering the withdrawal cases. All statistical analyses were performed using JMP ver. 15.0.0 software (SAS Institute, Cary, NC). The primary and secondary endpoints were analyzed using the paired t test. Comparisons of baseline characteristics between MRI-PDFF responders and non-responders were performed using the chi-squared test. A p value <0.05 was considered statistically significant.

Results

Patient characteristics

We screened patients who visited the Yokohama City University Hospital (Yokohama, Japan) between June 2018 and March 2019. The follow-up of participants ended in July 2019. Of the 23 patients (mean age = 50.6 ± 13.5 years) enrolled in this study, 22 completed the study protocol. Patient characteristics at baseline are shown in Table 1. Ten patients (43.5%) had type 2 diabetes and 17 (73.9%) had dyslipidemia. The mean baseline MRI-PDFF, MRE, ALT, and body mass index (BMI) were 18.4 ± 6.1%, 2.8 ± 1.2 kPa, 82.6 ± 50.9 U/L, and 28.3 ± 4.1 kg/m², respectively. Five patients took VitE orally, and all of them had been taking it for more than a year at the time they were enrolled.

Table 1.

Patient baseline characteristics (n = 23)

	Characteristics	Number (%) or mean ± SD
Demographics	Age, years	50.6 ± 13.5
	Sex, female/male	14/9 (60.9/39.1)
Comorbidities	Type 2 diabetes	10 (43.5%)
	Dyslipidemia	17 (73.9%)
	Hypertension	8 (34.8%)
Concomitant drug use	Anti-diabetic agents	8 (34.8%)
	Metformin	4 (17.4%)
	DPP4 inhibitor	1 (4.3%)
	Metformin + DPP4 inhibitor	1 (4.3%)
	DPP4 inhibitor + SGLT2 inhibitor	2 (8.7%)
	Anti-lipidemic drugs	15 (65.2%)
	Anti-hypertensives	6 (26.1%)
	Vitamin E	5 (21.7%)
	Ursodeoxycholic acid	1 (4.3%)

DPP4, dipeptidyl peptidase-4; SGLT2, sodium-glucose co-transporter 2.

Changes after the administration of VitB6

All 23 patients received oral VitB6 (pyridoxine hydrochloride 90 mg/day). One patient withdrew from the study because of rash; and therefore, 22 patients (9 men, 13 women, 50.1 ± 13.6 years) finished the protocol (see study flowchart, Fig. 1).

Fig. 1.

Flow chart of the study design. VitB6, vitamin B6; MRI, magnetic resonance imaging.

Patient characteristics before and after VitB6 treatment are shown in Table 2. After the administration of VitB6 for 12 weeks, serum ALT—the primary endpoint—did not change significantly (93.6 ± 46.9 to 93.9 ± 46.6, p = 0.976). On the other hand, MRI-PDFF, a parameter of liver fat accumulation, was significantly reduced (18.7 ± 6.1 to 16.4 ± 6.4, p<0.001; Fig. 2A) despite no significant changes in body weight (79.4 ± 15.8 to 79.3 ± 15.5, p = 0.978) or BMI (28.6 ± 4.1 to 28.4 ± 4.0, p = 0.922). Furthermore, MRI-PDFF was also significantly reduced even in those who did not take VitE (n = 17, 18.8 ± 6.9 to 16.7 ± 7.3, p = 0.0012). For hepatic fibrosis-related parameters, no significant changes were seen on MRE (Fig. 2B). There was no significant change in Fib-4 index, but type IV collagen 7s was significantly increased after VitB6 treatment. As for the lipid profile, no significant changes were found after VitB6 treatment, but HDL cholesterol tended to increase (37.4 ± 19.4 to 42.6 ± 22.5, p = 0.066). Regarding serum VitB6 levels, serum PL, a form of VitB6, was significantly increased (34.2 ± 77.2 to 224.3 ± 235.6, p<0.001), while serum PM and PN were increased, but not significantly.

Table 2.

Patient characteristics before and after Vitamin B6 treatment (n = 22)

	Characteristics	Before treatment	After treatment	p value
Liver image	MRI-PDFF†, %	18.7 ± 6.1	16.4 ± 6.4	<0.001**
	Magnetic resonance elastography, kPa	2.6 ± 0.92	2.5 ± 0.73	0.26
Metabolic factors	Weight, kg	79.4 ± 15.8	79.3 ± 15.5	0.978
	Body mass index, kg/m2	28.6 ± 4.1	28.4 ± 4.0	0.922
	Systolic blood pressure, mmHg	126.3 ± 4.4	124.5 ± 5.0	0.162
	Glucose, mg/dl	121.2 ± 29.8	119.8 ± 31.4	0.85
	Insulin, µU/ml	45.5 ± 58.2	45.0 ± 53.4	0.97
	HOMA-IR‡	10.5 ± 11.4	10.3 ± 12.0	0.936
	Hemoglobin A1c, %	6.2 ± 1.01	6.2 ± 1.09	0.93
Liver function	Platelet counts, ×109/L	218.1 ± 55.0	222.3 ± 57.9	0.449
	Alanine aminotransferase, IU/L	93.6 ± 46.9	93.9 ± 46.6	0.976
	Aspartate aminotransferase, IU/L	60.2 ± 29.6	58.0 ± 24.9	0.753
	γ-Glutamyl transferase, IU/L	90.0 ± 51.5	86.5 ± 37.8	0.567
	Alkaline phosphatase, IU/L	230.3 ± 72.9	232.5 ± 67.8	0.768
	Total bilirubin, mg/dl	0.83 ± 0.41	0.75 ± 0.30	0.131
Lipids	Total cholesterol, mg/dl	184.8 ± 35.4	193.3 ± 40.9	0.336
	Low-density lipoprotein cholesterol, mg/dl	104.3 ± 31.4	110.7 ± 37.9	0.394
	High-density lipoprotein cholesterol, mg/dl	37.4 ± 19.4	42.6 ± 22.5	0.066
	Triglyceride, mg/dl	229.4 ± 171.4	206.1 ± 131.4	0.302
Inflammation marker	High sensitivity C-reactive protein, mg/L	0.17 ± 0.16	0.12 ± 0.11	0.299
	Ferritin, ng/ml	189.6 ± 151.5	171.2 ± 107.4	0.243
Fibrosis marker	Type IV collagen 7s, ng/ml	5.2 ± 1.5	5.7 ± 1.9	0.020*
	Fibrosis-4 index	1.6 ± 1.0	1.6 ± 1.1	0.206
Renal function	Blood urea nitrogen, mg/dl	13.7 ± 4.9	13.2 ± 2.8	0.617
	Creatinine, mg/dl	0.80 ± 0.20	0.78 ± 0.21	0.164
Vitamin B6	Pyridoxamine, ng/ml	0.35 ± 0.24	0.87 ± 0.80	0.054
	Pyridoxal, ng/ml	34.2 ± 77.2	224.3 ± 235.6	<0.001***
	Pyridoxine, ng/ml	3.03 ± 0.13	4.61 ± 5.73	0.22
SF-8§ quality of life	Physical component	48.2 ± 7.6	48.3 ± 6.0	0.972
	Mental component	49.2 ± 7.7	49.5 ± 6.7	0.657

Data are expressed as number (%) or mean ± SD; *p<0.05, **p<0.01, ***p<0.001; ^†magnetic resonance imaging-based proton density fat fraction; ^‡homeostasis model assessment as an index of insulin resistance; ^§8-Item Short-Form Health Survey.

Fig. 2.

Change of MRI-PDFF and MRE. (A) Magnetic resonance imaging (MRI)-based proton density fat fraction (PDFF). (B) Magnetic resonance elastography (MRE) values before and after treatment with vitamin B6. *p<0.001.

Characteristics of MRI-PDFF responders

The 22 patients were divided into 2 groups based on MRI-PDFF findings as follows: MRI-PDFF responders (MRI-PDFF reduction >8%, n = 13) and MRI-PDFF non-responders (MRI-PDFF reduction <8%, n = 9). The baseline characteristics of the 2 groups are shown in Table 3. As for concomitant drugs, more patients received VitE supplementation in the responder group than in the non-responder group [n = 5 (38.5%) vs n = 0 (0%), p = 0.004]. And regarding fibrosis markers, the Fib-4 index in the MRI-PDFF responders at baseline was significantly lower than that in the non-responders (1.18 ± 0.85 vs 2.09 ± 0.97, p = 0.03). Also, type IV collagen 7s in the responders tended to be lower than in the non-responders (4.8 ± 1.7 vs 5.7 ± 0.9, p = 0.14). However, there was no significant difference between the two groups in the MRE, which is a more accurate parameter of liver fibrosis. MRI-PDFF and lipid profiles at baseline were not significantly different between the 2 groups.

Table 3.

Patient (n = 22) characteristics based on the reduction in hepatic fat accumulation

	Characteristics	MRI-PDFF† responders‡ (n = 13)	MRI-PDFF non-responders§ (n = 9)	p value
Demographics	Age, years	49.5 ± 15.8	51.0 ± 10.6	0.779
	Male	8 (61.5%)	5 (55.6%)	0.78
Comorbidities	Type 2 diabetes	5 (36.5%)	5 (55.6%)	0.429
	Dyslipidemia	9 (69.2%)	7 (77.8%)	0.658
	Hypertension	5 (38.5%)	3 (33.3%)	0.806
	Hyperuricemia	3 (23.0%)	0 (0%)	0.121
Concomitant drug use	Anti-diabetic agents	4 (23.1%)	5 (55.6%)	0.245
	Dipeptidyl peptidase-4 (DPP4)-inhibitor	4 (23.1%)	1 (11.1%)	0.279
	Metformin	2 (15.4%)	4 (44.4%)	0.132
	Sulfonylurea	0 (0%)	1 (11.1%)	0.219
	Anti-lipidemic agents	8 (61.5%)	7 (77.8%)	0.421
	Anti-hypertensives	4 (30.8%)	2 (22.2%)	0.658
	Vitamin E	5 (38.5%)	0 (0%)	0.004**
	Ursodeoxycholic acid	1 (7.7%)	0 (0%)	0.394
Liver image	MRI-PDFF, %	17.47 ± 1.67	20.37 ± 2.01	0.281
	Magnetic resonance elastography, kPa	2.71 ± 1.10	2.49 ± 0.61	0.602
Metabolic factors	Weight, kg	83.2 ± 19.0	74.8 ± 10.1	0.252
	Body mass index, kg/m2	29.2 ± 4.5	27.3 ± 2.5	0.336
	Systolic blood pressure, mmHg	125.5 ± 5.2	127.4 ± 3.0	0.335
	Glucose, mg/dl	116.8 ± 28.8	128.4 ± 31.9	0.403
	Insulin, µU/ml	41.2 ± 36.6	52.6 ± 85.4	0.676
	HOMA-IR¶	12.9 ± 13.4	7.8 ± 6.9	0.341
	Hemoglobin A1c, %	6.0 ± 1.0	6.4 ± 1.1	0.417
Liver function	Platelet counts	225.9 ± 49.6	203.6 ± 60.2	0.352
	Alanine aminotransferase, U/L	88.1 ± 45.9	92.6 ± 41.0	0.821
	Aspartate aminotransferase, U/L	64.3 ± 31.7	62.0 ± 25.6	0.864
	γ-glutamyl transferase, U/L	81.9 ± 49.2	94.3 ± 58.6	0.596
	Alkaline phosphatase, U/L	207.7 ± 70.5	261.2 ± 62.9	0.082
	Total bilirubin, mg/dl	0.82 ± 0.30	0.86 ± 0.54	0.824
Lipids	Total cholesterol, mg/dl	183.5 ± 42.3	185.4 ± 21.4	0.903
	Low-density lipoprotein cholesterol, mg/dl	102.5 ± 32.7	107.9 ± 29.2	0.694
	High-density lipoprotein cholesterol, mg/dl	33.8 ± 17.6	43.0 ± 20.5	0.281
	Triglyceride, mg/dl	257.2 ± 204.1	172.9 ± 93.5	0.263
Inflammation marker	High sensitivity C-reactive protein, mg/L	0.18 ± 0.20	0.17 ± 0.10	0.836
	Ferritin, ng/ml	201.5 ± 143.4	174.2 ± 161.5	0.681
Fibrosis marker	Type IV collagen 7s, ng/ml	4.8 ± 1.7	5.7 ± 0.9	0.147
	Fibrosis-4 index	1.18 ± 0.85	2.09 ± 0.97	0.030*
Renal function	Blood urea nitrogen, mg/dl	13.8 ± 3.1	13.4 ± 6.6	0.879
	Creatinine, mg/dl	0.84 ± 0.22	0.75 ± 0.17	0.261
Vitamin B6	Pyridoxamine, ng/ml	0.39 ± 0.28	0.27 ± 0.10	0.76
	Pyridoxal, ng/ml	45.7 ± 97.2	14.9 ± 25.2	0.36
	Pyridoxine, ng/ml	3.05 ± 0.17	3.00 ± 0.0	0.419
SF-8†† quality of life	Physical component	48.1 ± 6.9	48.5 ± 8.9	0.898
	Mental component	48.5 ± 8.7	50.2 ± 6.4	0.625

Data are expressed as number (%) or mean ± SD; *p<0.05, **p<0.01; ^†magnetic resonance imaging-based proton density fat fraction; reduction in lipid accumulation on MRI-PDFF ^‡>8% or ^§<8%; ^¶homeostasis model assessment as an index of insulin resistance; ^††8-Item Short-Form Health Survey.

Adverse events

Only one adverse event of rash occurred in 23 patients (4.34%); there were no serious adverse events.

Discussion

There are no known reports on the efficacy of VitB6 administration in patients with NAFLD, and this study investigates and provides the first known evidence that VitB6 administration ameliorates liver lipid deposition in patients with NAFLD.

Although the mechanism by which VitB6 ameliorates hepatic lipid accumulation has not been fully elucidated, one potential mechanism could be linked to homocysteine (Hcy) catabolism.

PLP, a physiologically active form of VitB6, functions as a coenzyme of cysteine-β-synthase (CBS) and cystathionine-γ-lyase (CGL).⁽¹⁹⁾ Because CBS and CGL contribute to Hcy catabolism, inadequate amounts of VitB6 result in the accumulation of Hcy. Hcy causes protein misfolding in the endoplasmic reticulum (ER), which leads to the ER stress response. ER stress induces the activation of the transcription factor sterol response element-binding protein 1c and causes de novo lipogenesis.⁽²⁰⁾ Thus, based on this mechanism, VitB6 deficiency is thought to induce hepatic lipid accumulation. In patients with NAFLD, plasma PLP levels are lower than those in healthy patients, while plasma Hcy levels are higher.^(14,21,22) It has also been reported that supplementation with B vitamins, including B6, lowered blood Hcy levels in subjects with one or more components of metabolic syndrome.⁽²³⁾ Therefore, VitB6 supplementation may reduce hepatic fat via the catabolism of Hcy.

Weight loss due to dietary changes or exercise has been reported to be associated with a reduction in hepatic fat.^(24–26) However, in the present study, there were no significant changes in body weight or BMI compared to pre-treatment, suggesting that the decrease in MRI-PDFF was not likely an effect of diet or exercise-induced weight loss. Also, there were no significant changes in metabolic factors such as blood lipids, blood glucose, or HOMA-IR, indicating that VitB6 supplementation may have a stronger effect on hepatic fat deposition than other parameters associated with metabolic syndrome.

Note that in this study, mean HDL cholesterol was elevated after VitB6 administration, although not significantly. The detailed mechanism is not fully elucidated, but a relationship between plasma VitB6 levels and dyslipidemia has been reported, and it is possible that VitB6 supplementation may have increased the level of plasma HDL cholesterol.⁽²⁷⁾ Also, the mean type IV collagen 7s level increased after VitB6 treatment. While the underlying mechanism remains unclear, it is possible that VitB6 is involved in type IV collagen 7s synthesis. Although this marker is used for the assessment of liver fibrosis, liver stiffness assessed by MRE tended to be decreased.

We also investigated patient factors associated with the MRI-PDFF response. The patients were divided into two groups based on the change in MRI-PDFF; we found that there were significantly more vitamin E (VitE) users in the responder group. All VitE users had been taking it for more than one year at the time of registration. Furthermore, even in those not taking VitE, MRI-PDFF was significantly reduced after 12 weeks. Therefore, it is unlikely that VitE alone improved hepatic fat accumulation. VitB6 alone appears to ameliorate liver fat, but VitB6 and VitE may have a synergistic effect.

This study has some limitations. First, our study was conducted in a single center, with a small sample size, and included a short treatment period of 12 weeks. Second, the study was a single-arm study without a control group. Third, pathological findings of the liver were not evaluated by liver biopsy. Future studies should be conducted in randomized control studies, with larger sample sizes, patients from multiple centers, and longer treatment periods.

VitB6 administration significantly ameliorates hepatic fat accumulation. Since VitB6 is an inexpensive agent with few side effects, it could be a potential novel therapeutic agent for the treatment of NAFLD. Long-term and large-scale randomized controlled trials are required to confirm the therapeutic effect of VitB6 in patients with NAFLD.

Author Contributions

TKobayashi and TKessoku have interpreted data and drafted the work. AO, MI, YH, YO, KI, MY, and SS have acquired and analyzed data. AN has revised the work. All authors have read and approved the final manuscript.

Abbreviations

ALT

alanine aminotransferase

AST

aspartate aminotransferase

BMI

body mass index

CBS

coenzyme of cysteine-β-synthase

CGL

cystathionine-γ-lyase

ER

endoplasmic reticulum

Fib-4

fibrosis-4

Hcy

homocysteine

HDL

high-density lipoprotein

HOMA-IR

homeostasis model assessment of insulin resistance

MRE

magnetic resonance elastography

MRI-PDFF

magnetic resonance imaging-based proton density fat fraction

NAFLD

nonalcoholic fatty liver disease

PL

pyridoxal

PLP

pyridoxal 5'-phosphate

PM

pyridoxamine

PMP

pyridoxamine 5'-phosphate

PN

pyridoxine

PNP

pyridoxine 5'-phosphate

SREBP1c

sterol response element-binding protein 1c

UMIN

University Hospital Medical Information Network

VitB6

vitamin B6

VitE

vitamin E

Conflict of Interest

No potential conflicts of interest were disclosed.