Abstract
Hyperuricemia reflects increased insulin resistance, and uric acid (UA) may serve as a predictive marker for the development of metabolic dysfunction-associated steatotic liver disease (MASLD); however, few studies have investigated this condition in the Japanese population. Thus, this retrospective observational study aimed to investigate the association of hyperuricemia with the risk of MASLD or metabolic and alcohol-related liver disease (MetALD) in individuals undergoing health checkups. A cross-sectional analysis was performed on 58,110 individuals, dividing them into quartile groups according to UA values for men and women (Q1 being the lowest and Q4 being the highest), and examining the complication rate of MASLD/MetALD. Subsequently, among 22,364 individuals without MASLD/MetALD, the relationship between UA at baseline and MASLD/MetALD development during follow-up was investigated using Cox proportional hazard models. In the cross-sectional analysis, the higher UA group had a higher complication rate of MASLD/MetALD in both men and women. In the follow-up analysis, both genders in the higher UA quartiles had a significantly higher incidence of MASLD/MetALD than those in the lower quartiles. Multivariate Cox proportional hazards analysis revealed that Q4 had a significantly higher hazard ratio than Q1 for both genders. These trends were the same in the time-dependent body mass index (BMI) model, which incorporated BMI as a time-dependent variable. High UA levels may serve as a predictive marker for MASLD/MetALD development. UA monitoring during health checkups could enable early detection and provision of intervention, improving patient outcomes.
Introduction
In recent years, the global prevalence of nonalcoholic fatty liver disease (NAFLD) has significantly increased. At present, it is the most common chronic liver disease [1-4]. NAFLD is associated with various extrahepatic complications, including cardiovascular disease, chronic kidney disease, and specific cancers. In addition, NAFLD imposes a significant economic burden and negatively affects the health-related quality of life [5]. As such, the increased incidence of NAFLD is a significant public health issue, and strategies for preventing and managing NAFLD are urgently needed. Notably, a recent study reported a significant increase in NAFLD among younger individuals aged <50 years [6]. Given these trends, younger individuals, particularly those undergoing health checkups, at high risk for NAFLD must be identified for early implementation of lifestyle interventions, including weight reduction, to prevent disease onset.
Even if patients present with multiple risk factors for NAFLD, such as obesity and type 2 diabetes, they are often excluded from clinical trials. This exclusion is simply attributed to their alcohol consumption that slightly exceeds the strict diagnostic criteria for NAFLD. However, alcohol-associated/related liver disease and NAFLD are increasingly recognized to share overlapping pathophysiological features [7]. Therefore, a more comprehensive term for steatotic liver diseases (SLDs) associated with metabolic dysfunction has been called for. In response to this exigency, the American Association for the Study of Liver Diseases and the European Association for the Study of the Liver, in collaboration with the Asociación Latinoamericana para el Estudio del Hígado, proposed the term metabolic dysfunction-associated SLD (MASLD), a new term for SLD and an alternative to NAFLD. MASLD was redefined to include the presence of at least one of the five cardiometabolic risk factors (CMRFs) based on its diagnostic criteria [8]. These CMRFs are all associated with insulin resistance, which is strongly linked to SLD development and progression [9, 10]. Furthermore, among patients with SLD meeting the MASLD criteria, those with high alcohol intake (weekly intake of 140–350 g for women, 210–420 g for men) are now classified as having MetALD [8].
Hyperuricemia acts as a precursor to gout and can be due to the overproduction or insufficient excretion of serum uric acid (UA). Furthermore, numerous studies have reported that hyperuricemia reflects high insulin resistance and is associated with the components of metabolic syndrome, including glucose metabolism disorders, abnormal lipid metabolism, obesity, and hypertension [11-14]. Given that insulin resistance is mainly involved in the onset and progression of NAFLD [15], UA levels may serve as a predictor for NAFLD development, and serum UA levels were reported to be associated with NAFLD [16]. Considering these findings, UA levels could be a prediction marker for MASLD onset; however, very few studies have investigated this in Japanese populations. Thus, this retrospective observational study was conducted using health check-up data to investigate the incidence of hyperuricemia and MASLD.
Materials and Methods
Study design and population
Participants who attended medical checkups at the Mirraza Shinjuku Tsurukame Clinic, Koganei Tsurukame Clinic, and Shinjuku Tsurukame Clinic (Tokyo, Japan) between May 1, 2015, and March 31, 2024, were assessed for potential inclusion in this retrospective observational study. The study included individuals aged 20 years or older who had received abdominal ultrasonography. For the cross-sectional analysis, only data from the earliest medical checkup were used if participants had more than one checkup during the study period. In the follow-up analysis, individuals without MASLD/MetALD at their initial visit were tracked to determine whether they developed these conditions upon subsequent checkups, including abdominal ultrasonography. None of the participating clinics provided health guidance to those diagnosed with MASLD/MetALD.
In our study, we excluded the following individuals. First, those with a history of using uric acid-lowering agents were excluded. Second, individuals with an FIB-4 index >2.67, calculated as age (years) × aspartate aminotransferase (AST; IU/L)/(platelet count [109/L] × (alanine aminotransferase [ALT; U/L])1/2) [17], which suggests suspected advanced-stage liver fibrosis, were also excluded. This exclusion was made because patients with advanced liver fibrosis may experience fluctuations in serum uric acid levels due to changes in circulating blood volume, which could complicate the evaluation of the relationship between uric acid levels and the development of fatty liver disease. Moreover, advanced liver fibrosis is frequently associated with impaired renal function, a condition that may contribute to hyperuricemia. In addition, severe fibrosis of the liver can lead to malnutrition, which may, in turn, result in hypouricemia. Given these, which could interfere with accurately assessing the association between uric acid and fatty liver disease, we excluded individuals with advanced liver fibrosis, identified as those with an FIB-4 index >2.67. Third, those with excessive alcohol intake (women who consume >350 g of alcohol per week and men who consume >420 g of alcohol per week), those with a history of viral hepatitis, and those with missing data were excluded. In total, 58,110 participants were included in the cross-sectional analysis and 22,364 in the follow-up analysis (Fig. 1).
Fig. 1
Flow chart of the study population.
This study adhered to the principles outlined in the Declaration of Helsinki. The ethics committee of Shinjuku Tsurukame Clinic approved the study protocol (Approval no. 2402). Due to its retrospective observational design, obtaining written informed consent was deemed unnecessary. Instead, patients were given the option to opt out, with information about the analysis disclosed on the hospital’s website.
Collection of clinical data
The methods used to collect clinical data, including blood tests and lifestyle-related factors, were consistent with those detailed in our previous study [18]. Smoking status was categorized into current smokers and nonsmokers. An exercise habit was defined as engaging in exercise for at least 30 minutes per session, at least two days per week, for a minimum duration of one year. Physical activity was defined as walking or performing an equivalent activity for at least one hour daily. Fast walking referred to walking at a pace faster than individuals of similar age and sex. Fast eating was defined as eating at a speed faster than others. Additionally, meal just before bedtime was defined as eating within two hours before sleeping at least three times per week. Absence of breakfast was defined as not eating breakfast at least three times per week. Good sleeping was defined as waking up feeling well-rested. Regarding alcohol consumption, drinking frequency was categorized into daily, occasional, or rare drinking. The amount of alcohol consumed in one drinking session was standardized as approximately 20 g of alcohol per drink and classified into four categories: less than 1 drink, 1–2 drinks, 2–3 drinks, and 3 or more drinks. Weekly alcohol consumption was estimated using the following calculation: 1 drink = 20 g, 1–2 drinks = 30 g, 2–3 drinks = 50 g, and 3 or more drinks = 70 g, multiplied by 7 for daily drinkers and 2 for occasional drinkers. For participants who reported rare drinking, alcohol consumption was assumed to be 0 g per week.
UA levels were measured to one decimal place, and the overall distribution of UA levels for men and women was divided into quartiles (Q1–Q4). For men, Q1, Q2, Q3, and Q4 were defined as <5.4, 5.4–6.1, 6.2–6.9, and ≥7.0 mg/dL, respectively. For women, the quartile values were <3.8, 3.8–4.3, 4.4–4.9, and ≥5.0 mg/dL, respectively. The fatty liver index (FLI) was calculated as (e0.953 × loge (TG [mg/dL]) + 0.139 × BMI + 0.718 × loge (γ-GTP [U/L]) + 0.053 × WC [cm] – 15.745)/(1 + e0.953 × loge (TG) + 0.139 × BMI + 0.718 × loge (γ-GTP [U/L]) + 0.053 × WC [cm] – 15.745) × 100 [19].
Ultrasound finding analysis
Abdominal ultrasonography was performed by a trained clinical laboratory technologist using the Aplio MX, Aplio 300, Aplio a450, or Aplio a550 ultrasound systems (Canon Medical Systems, Tokyo, Japan). Steatotic liver was diagnosed based on one of the following criteria, as previously described [18]: 1) bright liver with high intensity, 2) hepatorenal contrast, 3) obscured blood vessels, or 4) deep attenuation in the liver. The initial diagnosis of steatotic liver was made by a clinical laboratory technologist using abdominal ultrasonography, and the final diagnosis was confirmed by a board-certified gastroenterologist from the Japanese Gastroenterological Association after reviewing the ultrasound images.
Definition of MASLD and MetALD
MASLD was diagnosed based on the consensus statement [8]. Participants were classified as having MASLD if they exhibited a steatotic liver on abdominal ultrasonography, consumed less than 140 g of alcohol per week for women or less than 210 g per week for men, and met at least one of the following five CMRF criteria: 1) a body mass index (BMI) of ≥23 kg/m2 or a waist circumference (WC) of ≥85 cm for men or ≥90 cm for women [20], 2) a fasting plasma glucose (FPG) level of ≥100 mg/dL or the use of medications for type 2 diabetes, 3) blood pressure ≥130/85 mmHg or the use of antihypertensive medications, 4) triglyceride (TG) levels ≥150 mg/dL or the use of lipid-lowering medications, or 5) high-density lipoprotein (HDL) cholesterol levels ≤40 mg/dL for men or ≤50 mg/dL for women or the use of lipid-lowering drugs. MetALD was defined for participants with steatotic liver observed on abdominal ultrasonography, alcohol consumption ranging from 140–350 g per week for women or 210–420 g per week for men, and at least one of the aforementioned CMRF criteria. The endpoint for the follow-up analysis was the occurrence of MASLD or MetALD at the time of the second or any subsequent health check-up.
Statistical analysis
Data are presented as means ± standard deviations, median with interquartile range, or percentage, in accordance with the distribution characteristics of the data. Analysis of variance was used to test continuous variables, and the χ2-test was used to analyze categorical variables.
Receiver operating characteristic (ROC) analyses were performed to determine the areas under the curve (AUC) for UA levels to assess their discriminatory power for identifying MASLD/MetALD. Optimal cutoff values were determined using Youden’s index [21].
The Cox proportional hazards model was used to investigate whether Q2, Q3, and Q4 patients had a higher incidence of MASLD/MetALD than those in Q1 patients who were not diagnosed with MASLD/MetALD at the initial visit. The variables included in the multivariate Cox proportional hazard model were selected through a stepwise approach. Logarithmic transformations were applied to variables such as gamma-glutamyl transferase (γ-GTP), TG, FLI, and alcohol consumption to normalize their distributions, using the natural logarithm (base e). For participants without alcohol consumption, the value was log-transformed as 1 g/week. Furthermore, considering that BMI may change during the follow-up period and that such changes could significantly affect MASLD/MetALD development [18], we developed a time-dependent BMI model that includes BMI as a time-dependent variable instead of BMI at the initial visit.
P-values <0.05 were deemed significant. All statistical analyses were performed using IBM SPSS Statistics for Windows version 21.0 (IBM Corp., Armonk, NY, USA).
Results
This study enrolled 58,110 participants with a mean age of 45.4 ± 9.7 years, and approximately 53% of them were men. Table 1 presents clinical characteristics, comorbidities, laboratory data, and lifestyle-related factors in the cross-sectional analysis. BMI and WC increased progressively from Q1 to Q4 in men and women. The proportion of patients with MASLD/MetALD also increased progressively from Q1 to Q4 in both sexes. All laboratory data, including FPG, hemoglobin A1c (HbA1c), TG, HDL cholesterol, low-density lipoprotein (LDL)-cholesterol, AST, ALT, γ-GTP, UA, creatinine (Cr), estimated glomerular filtration rate (eGFR), C-reactive protein, white blood cell, hemoglobin (Hb), platelet count (Plt), systolic blood pressure, diastolic blood pressure (DBP), FIB-4 index, and FLI, showed significant differences across from Q1 to Q4. Regarding the lifestyle-related factors across the UA quartiles (Q1–Q4), smoking prevalence increased progressively with higher quartiles, with the highest smoking rates observed in Q4 in both sexes. Exercise habits were more prevalent in men within lower quartiles, whereas women showed a higher prevalence of exercise in higher quartiles. The proportion of physically active individuals was greater among men in lower quartiles, whereas no significant difference was found across quartiles in women. Non-intake of breakfast increased progressively in higher quartiles for both sexes. The proportion of daily drinking and weekly alcohol consumption increased progressively in higher quartiles for both sexes.
Table 1 Patient characteristics stratified by uric acid quartiles for both genders.
|
All participants |
men |
women |
| Q1 |
Q2 |
Q3 |
Q4 |
p |
Q1 |
Q2 |
Q3 |
Q4 |
p |
| n |
58,110 |
6,949 |
7,955 |
7,783 |
8,377 |
|
6,094 |
6,848 |
6,465 |
7,639 |
|
| Range of UA (mg/dL) |
|
<5.4 |
5.4–6.1 |
6.2–6.9 |
≥7.0 |
|
<3.8 |
3.8–4.3 |
4.4–4.9 |
≥5.0 |
|
| Age (years) |
45.4 ± 9.7 |
46.4 ± 10.4 |
45.7 ± 10.0 |
45.4 ± 9.8 |
45.0 ± 9.7 |
<0.001 |
43.5 ± 8.4 |
44.2 ± 9.0 |
45.0 ± 9.6 |
47.5 ± 10.1 |
<0.001 |
| Body mass index (kg/m2) |
22.7 ± 3.6 |
22.7 ± 3.0 |
23.1 ± 3.2 |
23.8 ± 3.2 |
25.0 ± 3.6 |
<0.001 |
20.4 ± 2.6 |
20.9 ± 2.8 |
21.4 ± 3.1 |
23.1 ± 4.3 |
<0.001 |
| Waist circumference (cm) |
81.2 ± 11.9 |
82.0 ± 8.7 |
83.4 ± 8.7 |
85.3 ± 8.8 |
88.7 ± 9.6 |
<0.001 |
72.4 ± 13.7 |
75.2 ± 9.9 |
77.2 ± 9.4 |
81.2 ± 11.1 |
<0.001 |
| Comorbidities |
| Hypertension (n, [%]) |
4,269 (7) |
650 (9) |
727 (9) |
786 (10) |
868 (10) |
0.026 |
133 (2) |
186 (3) |
238 (4) |
681 (9) |
<0.001 |
| Diabetes (n, [%]) |
1,179 (2) |
326 (5) |
268 (3) |
187 (2) |
158 (2) |
<0.001 |
49 (1) |
48 (1) |
36 (1) |
107 (1) |
<0.001 |
| Dyslipidemia (n, [%]) |
3,207 (6) |
496 (7) |
558 (7) |
495 (6) |
451 (5) |
<0.001 |
153 (3) |
229 (3) |
253 (4) |
572 (7) |
<0.001 |
| MASLD/MetALD (n, [%]) |
16,537 (29) |
1,850 (27) |
2,567 (32) |
3,222 (42) |
4,835 (58) |
<0.001 |
341 (6) |
578 (8) |
833 (13) |
2,300 (30) |
<0.001 |
| Laboratory data |
| FPG
(mg/dL) |
87.6 ± 14.2 |
91.0 ± 19.4 |
90.1 ± 16.9 |
89.9 ± 14.9 |
90.3 ± 13.9 |
<0.001 |
83.4 ± 9.7 |
83.6 ± 9.3 |
84.2 ± 9.5 |
86.3 ± 12.0 |
<0.001 |
| HbA1c
(%) |
5.4 ± 0.5
(n =
54,214) |
5.5 ± 0.7 |
5.5 ± 0.6 |
5.5 ± 0.5 |
5.5 ± 0.5 |
<0.001 |
5.3 ± 0.3 |
5.3 ± 0.4 |
5.4 ± 0.3 |
5.4 ± 0.5 |
<0.001 |
| TG
(mg/dL) |
99.4 ± 77.9 |
99.5 ± 69.8 |
105.4 ± 72.7 |
119.8 ± 99.6 |
149.5 ± 111.2 |
<0.001 |
65.7 ± 33.4 |
68.5 ± 36.9 |
74.0 ± 38.4 |
93.1 ± 60.0 |
<0.001 |
| HDL-Cholesterol (mg/dL) |
64.2 ± 17.4 |
60.0 ± 15.0 |
58.8 ± 15.1 |
57.0 ± 14.8 |
54.2 ± 14.3 |
<0.001 |
73.8 ± 15.4 |
73.1 ± 15.6 |
72.4 ± 16.5 |
69.5 ± 18.3 |
<0.001 |
| LDL-Cholesterol (mg/dL) |
121.6 ± 31.5 |
119.9 ± 29.7 |
123.1 ± 30.2 |
126.6 ± 30.9 |
132.0 ± 33.0 |
<0.001 |
111.1 ± 28.0 |
114.2 ± 28.9 |
117.1 ± 30.2 |
124.2 ± 34.1 |
<0.001 |
| AST
(IU/L) |
22.1 ± 16.9 |
21.8 ± 8.9 |
22.7 ± 13.4 |
24.0 ± 11.0 |
27.7 ± 36.5 |
<0.001 |
18.3 ± 6.1 |
18.8 ± 6.5 |
19.4 ± 6.6 |
21.8 ± 12.6 |
<0.001 |
| ALT
(IU/L) |
22.3 ± 21.4 |
22.8 ± 15.1 |
24.7 ± 21.7 |
27.3 ± 19.0 |
34.9 ± 35.2 |
<0.001 |
14.0 ± 8.4 |
14.7 ± 9.4 |
15.8 ± 9.7 |
19.8 ± 20.1 |
<0.001 |
| γ-GTP
(IU/L) |
35.1 ± 47.2 |
35.9 ± 36.9 |
39.9 ± 48.1 |
46.2 ± 61.9 |
60.9 ± 71.7 |
<0.001 |
17.8 ± 16.0 |
18.8 ± 14.1 |
22.1 ± 21.4 |
29.3 ± 38.8 |
<0.001 |
| UA
(mg/dL) |
5.4 ± 1.4 |
4.6 ± 0.7 |
5.8 ± 0.2 |
6.5 ± 0.2 |
7.8 ± 0.7 |
<0.001 |
3.2 ± 0.4 |
4.1 ± 0.2 |
4.6 ± 0.2 |
5.7 ± 0.7 |
<0.001 |
| Cr
(mg/dL) |
0.8 ± 0.2 |
0.8 ± 0.3 |
0.9 ± 0.2 |
0.9 ± 0.2 |
0.9 ± 0.2 |
<0.001 |
0.6 ± 0.1 |
0.6 ± 0.1 |
0.6 ± 0.1 |
0.7 ± 0.1 |
<0.001 |
| eGFR
(mL/min/1.73 m2) |
80.1 ± 14.2 |
81.4 ± 13.8 |
79.5 ± 12.7 |
77.6 ± 12.6 |
75.1 ± 13.1 |
<0.001 |
87.9 ± 15.7 |
83.8 ± 14.2 |
81.2 ± 13.9 |
77.0 ± 14.3 |
<0.001 |
| CRP
(mg/dL) |
0.10 ± 0.31
(n =
21,783) |
0.11 ± 0.48 |
0.10 ± 0.41 |
0.10 ± 0.28 |
0.13 ± 0.30 |
0.017 |
0.06 ± 0.24 |
0.07 ± 0.21 |
0.07 ± 0.18 |
0.12 ± 0.27 |
<0.001 |
| WBC
(number/μL) |
5,678 ± 1,775 |
5,566 ± 1,573 |
5,652 ± 1,528 |
5,794 ± 1,512 |
6,071 ± 2,794 |
<0.001 |
5,394 ± 1,514 |
5,430 ± 1,440 |
5,539 ± 1,451 |
5,822 ± 1,607 |
<0.001 |
| Hb
(g/dL) |
14.2 ± 1.5 |
15.0 ± 1.0 |
15.1 ± 1.0 |
15.2 ± 1.0 |
15.4 ± 1.0 |
<0.001 |
12.8 ± 1.3 |
13.0 ± 1.1 |
13.2 ± 1.1 |
13.4 ± 1.1 |
<0.001 |
| Plt
(104/μL) |
26.3 ± 5.8 |
25.4 ± 5.3 |
25.5 ± 5.5 |
25.8 ± 5.5 |
26.1 ± 5.6 |
<0.001 |
26.7 ± 6.2 |
26.7 ± 6.0 |
27.1 ± 6.1 |
27.7 ± 6.3 |
<0.001 |
| SBP
(mmHg) |
116.0 ± 16.6 |
116.9 ± 15.5 |
117.7 ± 15.2 |
119.8 ± 15.5 |
123.0 ± 16.0 |
<0.001 |
108.4 ± 14.5 |
109.8 ± 15.3 |
112.0 ± 16.4 |
117.0 ± 18.1 |
<0.001 |
| DBP
(mmHg) |
72.8 ± 12.3 |
73.8 ± 11.7 |
74.5 ± 11.5 |
76.3 ± 12.0 |
78.9 ± 12.3 |
<0.001 |
66.4 ± 10.7 |
67.7 ± 10.9 |
69.1 ± 11.4 |
72.7 ± 12.4 |
<0.001 |
| FIB-4
index |
0.9 ± 0.4 |
0.9 ± 0.4 |
0.9 ± 0.5 |
0.9 ± 0.4 |
0.9 ± 0.5 |
<0.001 |
0.9 ± 0.4 |
0.9 ± 0.4 |
0.9 ± 0.4 |
0.9 ± 0.5 |
<0.001 |
| FLI |
23.1 ± 24.2 |
22.8 ± 21.6 |
26.2 ± 22.5 |
32.6 ± 24.4 |
45.4 ± 27.4 |
<0.001 |
6.6 ± 9.3 |
8.3 ± 11.1 |
11.5 ± 14.6 |
22.1 ± 23.6 |
<0.001 |
| lifestyle-related factors |
| Current Smoker (n, [%]) |
10,483 (18) |
1,827 (26) |
2,050 (26) |
2,918 (26) |
2,326 (28) |
0.015 |
418 (7) |
512 (7) |
541 (8) |
791 (10) |
<0.001 |
| Exercise habit (n, [%]) |
12,728 (22) |
1,922 (28) |
2,053 (26) |
2,047 (26) |
1,939 (23) |
<0.001 |
953 (16) |
1,154 (17) |
1,184 (18) |
1476 (19) |
<0.001 |
| Physically Active (n, [%]) |
25,889 (45) |
3,238 (47) |
3,647 (46) |
3,471 (45) |
3,548 (42) |
<0.001 |
2,707 (44) |
3,017 (44) |
2,863 (44) |
3,398 (44) |
0.960 |
| Fast
walking (n, [%]) |
31,079 (54) |
4,065 (58) |
4,734 (60) |
4,488 (58) |
4,783 (57) |
0.001 |
2,887 (47) |
3,258 (48) |
3,185 (49) |
3,724 (49) |
0.088 |
| Fast
eating (n, [%]) |
34,261 (59) |
3,943 (57) |
4,344 (55) |
4,009 (52) |
4,051 (48) |
<0.001 |
4,125 (68) |
4,590 (67) |
4,283 (66) |
4,916 (64) |
<0.001 |
| Meal
just before bedtime (n, [%]) |
22,172 (38) |
3,069 (44) |
3,647 (46) |
3,657 (47) |
4,112 (49) |
<0.001 |
1,604 (26) |
1,869 (27) |
1,887 (29) |
2,327 (30) |
<0.001 |
| Absence of breakfast (n, [%]) |
15,165 (26) |
1,894 (27) |
2,296 (29) |
2,391 (31) |
2,910 (35) |
<0.001 |
1,168 (19) |
1,355 (20) |
1,378 (21) |
1,773 (23) |
<0.001 |
| Good
sleeping (n, [%]) |
36,877 (64) |
4,442 (64) |
5,134 (65) |
5,107 (66) |
5,378 (64) |
0.138 |
3,836 (63) |
4,232 (62) |
4,087 (63) |
4,661 (61) |
0.026 |
| Daily
drinking (n, [%]) |
16,606 (29) |
2,303 (33) |
2,892 (36) |
3,282 (42) |
3,913 (47) |
<0.001 |
670 (11) |
836 (12) |
1,022 (16) |
1,688 (22) |
<0.001 |
| Alcohol consumption (g/week) |
84.7 ± 111.0 |
95.4 ± 115.5 |
103.8 ± 119.4 |
117.9 ± 126.9 |
130.2 ± 133.3 |
<0.001 |
40.6 ± 67.1 |
44.4 ± 68.3 |
53.5 ± 81.0 |
69.5 ± 97.9 |
<0.001 |
Data are shown as mean ± SD or number (percentage). p value represents the difference among the groups stratified by uric acid quartiles for both genders in means (ANOVA), or percent (chi square test).
ALT, alanine aminotransferase; AST, aspartate aminotransferase; CRP, C-reactive protein; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; FLI, fatty liver index; FPG, fasting plasma glucose; γ-GTP, gamma-glutamyl transferase; Hb, hemoglobin; HbA1c, hemoglobin A1c; HDL-Cholesterol, high-density lipoprotein cholesterol; LDL-Cholesterol, low-density lipoprotein cholesterol; MASLD, metabolic dysfunction-associated steatotic liver disease; MetALD, metabolic and alcohol related/associated liver disease; Plt, platelet count; SBP, systolic blood pressure; TG, triglyceride; UA, uric acid; WBC, white blood cell count.
Table 2 shows the Pearson correlation coefficients between UA levels and other parameters. All variables, except for HbA1c in men, showed a significant association with UA levels.
Table 2 The Pearson correlation coefficients between uric acid levels and other parameters.
|
All participants |
Men |
Women |
| Age, y |
0.037** |
–0.058** |
0.162** |
| Body mass index, kg/m2 |
0.408** |
0.276** |
0.328** |
| Waist circumference, cm |
0.435** |
0.274** |
0.325** |
| FPG, mg/dL |
0.146** |
–0.015* |
0.116** |
| HbA1c, % |
0.094** |
–0.003 |
0.148** |
| TG, mg/dLa |
0.427** |
0.270** |
0.280** |
| HDL, mg/dL |
–0.355** |
–0.149** |
–0.314** |
| LDL, mg/dL |
0.214** |
0.155** |
0.173** |
| AST, IU/L |
0.293** |
0.209** |
0.200** |
| ALT, IU/L |
0.353** |
0.231** |
0.203** |
| γ-GTP, IU/La |
0.475** |
0.266** |
0.302** |
| Cr, mg/dL |
0.424** |
0.110** |
0.249** |
| eGFR, mL/min/1.73 m2 |
–0.257** |
–0.180** |
–0.275** |
| CRP, mg/dL |
0.070** |
0.040** |
0.104** |
| WBC, /μL |
0.125** |
0.104** |
0.125** |
| Hb, g/dL |
0.520** |
0.136** |
0.164** |
| Plt, 104/μL |
–0.023** |
0.055** |
0.079** |
| SBP, mmHg |
0.277** |
0.160** |
0.212** |
| DBP, mmHg |
0.313** |
0.172** |
0.217** |
| FIB-4 index |
0.010* |
–0.036** |
0.061** |
| FLIa |
0.555** |
0.348** |
0.411** |
| Alcohol consumption, g/weeka |
0.233** |
0.096** |
0.112** |
ALT, alanine aminotransferase; AST, aspartate aminotransferase; CRP, C-reactive protein; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; FLI, fatty liver index; FPG, fasting plasma glucose; γ-GTP, gamma-glutamyl transferase; Hb, hemoglobin; HbA1c, hemoglobin A1c; HDL-Cholesterol, high-density lipoprotein cholesterol; LDL-Cholesterol, low-density lipoprotein cholesterol; MASLD, metabolic dysfunction-associated steatotic liver disease; MetALD, metabolic and alcohol related/associated liver disease; Plt, platelet count; SBP, systolic blood pressure; TG, triglyceride; UA, uric acid; WBC, white blood cell count.
*p < 0.05, **p < 0.001
a log-transformed
Fig. 2 shows the ROC curves for predicting MASLD/MetALD development based on UA levels in all participants (Fig. 2A), men (Fig. 2B), and women (Fig. 2C), with AUC values of 0.735, 0.646, and 0.724 and optimal cutoff values for UA levels of 5.65, 6.35, and 4.75 mg/dL, respectively. Then, participants were categorized into MASLD and MetALD groups based on their weekly alcohol consumption, and ROC curves were created for each group. Women who consumed 140–350 g of alcohol per week and men who consumed 210–420 g per week were classified into the MetALD group, whereas women who consumed <140 g and men who consumed <210 g were classified into the MASLD group. The results of this analysis are presented in Supplementary Fig. 1. For the MASLD group, the AUC values were 0.752, 0.664, and 0.731 for all participants, men, and women, with optimal cutoff values of 5.45, 6.25, and 4.75 mg/dL, respectively. For the MetALD group, the AUC values were 0.662, 0.600, and 0.671 for all participants, men, and women, with optimal cutoff values of 6.05, 6.45, and 5.15 mg/dL, respectively.
Fig. 2
ROC curves for predicting MASLD/MetALD based on uric acid levels. All participants (A), men (B), and women (C).
Of the 58,110 participants in the cross-sectional analysis, we examined whether 22,364 individuals who did not have MASLD/MetALD at the initial visit had developed MASLD/MetALD at least once during the follow-up period. During the average follow-up period of 3.46 ± 2.42 and 3.73 ± 2.50 years in men and women, respectively, MASLD/MetALD developed in 2,587 men and 1,374 women, respectively. Fig. 3 shows the Kaplan–Meier curve of the time to MASLD/MetALD development according to the quartiles of the UA levels in both sexes. In the overall analysis for both men (Fig. 3A) and women (Fig. 3B), the incidence of MASLD/MetALD was significantly higher in quartiles with higher UA levels than in lower quartiles. In the analysis of 4,306 men and 7,606 women without CMRFs at the initial visit, the incidence of MASLD/MetALD was significantly higher in quartiles with higher UA levels among men (Fig. 3C), whereas among women, only Q4 showed a significantly higher incidence than the others (Fig. 3D). Supplementary Figs. 2, 3 show the Kaplan–Meier curves for the incidence of MASLD/MetALD during the follow-up period according to the UA quartiles in each participant group with 1, 2, 3, or ≥4 CMRFs, for men and women, respectively. In men, a significant increase in the incidence of MASLD/MetALD was observed in UA quartiles of patients with one and two CMRF. In women, a significant increase in the incidence of MASLD/MetALD was observed only in the group with one CMRF as the UA quartiles increased.
Fig. 3
Kaplan–Meier curve of the time to the development for MASLD/MetALD according to the quartiles of the UA levels in both men and women. All men (A), all women (B), men without CMRF (C), and women without CMRF (D). CMRF, cardiometabolic risk factor.
Table 3 presents the results of the Cox proportional hazards model that examined the association between UA quartiles and MASLD/MetALD incidence in men and women. In the univariate analysis, higher UA quartiles were associated with a significantly increased risk of MASLD/MetALD in both sexes, with Q4 showing an HR of 1.831 (95% CI 1.632–2.054) in men and 2.204 (95% CI 1.184–2.580) in women. In the multivariate analysis, Q2–Q4 in men had a significantly higher HR than Q1, whereas in women, only Q4 had a significantly higher HR than Q1, and these trends remained unchanged in the time-dependent BMI model, where BMI was incorporated as a time-dependent variable.
Table 3 Univariate and multivariate Cox proportional hazards models examining the association between baseline uric acid quartiles and the incidence of MASLD/MetALD during follow-up.
|
Men (n = 10,451) |
Women (n = 11,913) |
| HR (95%CI) |
p |
HR (95%CI) |
p |
| Endpoint, n |
2,587 |
|
1,374 |
|
| Univariate |
| Q1 |
Reference |
|
Reference |
|
| Q2 |
1.205 (1.075–1.350) |
<0.001 |
1.180 (0.994–1.400) |
0.058 |
| Q3 |
1.487 (1.329–1.664) |
<0.001 |
1.494 (1.266–1.762) |
<0.001 |
| Q4 |
1.831 (1.632–2.054) |
<0.001 |
2.204 (1.184–2.580) |
<0.001 |
| Multivariate |
| Q1 |
Reference |
|
Reference |
|
| Q2 |
1.130 (1.008–1.267) |
0.037 |
1.077 (0.906–1.281) |
0.399 |
| Q3 |
1.313 (1.170–1.473) |
<0.001 |
1.138 (0.959–1.351) |
0.140 |
| Q4 |
1.386 (1.226–1.568) |
<0.001 |
1.351 (1.139–1.604) |
<0.001 |
| Age |
1.006 (1.002–1.011) |
0.008 |
1.016 (1008–1.023) |
<0.001 |
| BMI |
1.113 (1.083–1.144) |
<0.001 |
1.196 (1.169–1.224) |
<0.001 |
| Waist
circumference |
1.017 (1.007–1.027) |
0.001 |
1.016 (1.008–1.023) |
<0.001 |
| TG |
1.001 (1.001–1.002) |
<0.001 |
1.004 (1.003–1.005) |
<0.001 |
| HDL-Cholesterol |
0.989 (0.986–0.992) |
<0.001 |
0.990 (0.986–0.994) |
<0.001 |
| LDL-Cholesterol |
1.002 (1.001–1.002) |
<0.001 |
1.005 (1.003–1.011) |
<0.001 |
| AST |
0.986 (0.979–0.994) |
<0.001 |
NS |
|
| ALT |
1.015 (1.011–1.020) |
<0.001 |
NS |
|
| γ-GTP |
1.001 (1.000–1.002) |
0.012 |
NS |
|
| eGFR |
1.008 (1.005–1.012) |
<0.001 |
1.007 (1.003–1.011) |
0.001 |
| Plt |
1.012 (1.005–1.019) |
<0.001 |
1.012 (1.003–1.020) |
0.006 |
| SBP |
NS |
|
1.009 (1.005–1.012) |
<0.001 |
| Absence of breakfast |
1.179 (1.080–1.286) |
<0.001 |
NS |
|
| Exercise habit |
NS |
|
0.836 (0.718–0.974) |
0.021 |
| Time-dependent BMI model |
| Q1 |
Reference |
|
Reference |
|
| Q2 |
1.127 (1.013–1.288) |
0.025 |
1.080 (0.908–1.284) |
0.383 |
| Q3 |
1.304 (1.162–1.463) |
<0.001 |
1.141 (0.962–1.355) |
0.130 |
| Q4 |
1.377 (1.217–1.557) |
<0.001 |
1.339 (1.128–1.590) |
<0.001 |
ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; CI, confidence interval; eGFR, estimated glomerular filtration rate; γ-GTP, gamma-glutamyl transferase; HDL-Cholesterol, high-density lipoprotein cholesterol; HR, hazard ratio; LDL-Cholesterol, low-density lipoprotein cholesterol; NS, not selected; Plt, platelet count; SBP, systolic blood pressure; TG, triglyceride; WBC, white blood cell count.
Multivariate models were adjusted for variables selected by stepwise selection from the following: age, body mass index, waist circumference, fasting plasma glucose, triglycerides, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, aspartate aminotransferase, alanine aminotransferase, gamma-glutamyl transferase, estimated glomerular filtration rate, white blood cells, hemoglobin, platelets, systolic blood pressure, and life related factors including current smoker, exercise habit, physically active, fast walking, fast eating, meal just before bedtime, absence of breakfast, good sleeping, alcohol consumption.
Time-dependent BMI models are multivariate models for both genders in which BMI is included as a time-dependent variable instead of baseline BMI.
In Fig. 4, a sensitivity analysis was performed to assess the HRs of the initial serum UA level quartiles for MASLD development during the follow-up period, across different patient groups and stratified by sex. In men, participants aged <40 and 40 to <60 in Q4 had higher HRs than those in Q1. This was also observed in participants with a BMI <23 and 23 to <25. Furthermore, in participants with 0, 1, or 2 CMRFs, Q4 had significantly higher HRs than Q1. In addition, in the analysis by weekly alcohol consumption, Q4 consistently had a significantly higher HR than Q1. In women, participants aged <40 and 40 to <60, those with a BMI <23, and those in the CMRF 0 group had significantly higher HRs in Q4 than in Q1. In the group with weekly alcohol consumption of <140 g, Q4 had a significantly higher HR than Q1; however, in the group with weekly alcohol consumption of 140–350 g, even Q4 did not have a higher HR than Q1.
Fig. 4
Sensitivity analysis of Cox proportional hazards models examining the association of initial serum uric acid quartiles with the development of MASLD/MetALD stratified by gender, age, BMI, CMRF, and alcohol consumption. BMI, body mass index; CMRF, cardiometabolic risk factor.
Discussion
In this study, we first demonstrated a significant association between UA levels and MASLD development in a cross-sectional analysis of 58,110 participants. Furthermore, analysis of follow-up data from 22,364 individuals without MASLD showed that high UA levels were associated with an increased risk of MASLD development. This association remained significant even after adjusting for variables including BMI, WC, blood glucose levels, TG, HDL cholesterol, and lifestyle-related factors. Notably, even among participants without any CMRFs, those in the highest quartile (Q4) of UA levels had a significantly higher risk of MASLD development than those in the lowest quartile (Q1) for both sexes.
In the cross-sectional analysis, exercise habits were more prevalent in men within lower quartiles than in women who exhibited a higher prevalence of exercise in higher quartiles. Our data consistently revealed that men and women with exercise habits tended to be older (data not shown). Furthermore, considering that in men, higher quartiles of UA were associated with younger age, and in women, higher quartiles of UA were associated with older age, the inconsistent results between uric acid quartiles and exercise habits by sex mentioned earlier were thought to be influenced by age. This finding might be due to the fact that hyperuricemia is more strongly influenced by lifestyle and obesity in men, whereas hyperuricemia is more strongly influenced by the menopause-induced decrease in estrogen in women [22]. However, because this study did not obtain information related to menstrual cycles, further research is needed to investigate the factors contributing to hyperuricemia by sex.
Most MASLD and NAFLD cases are believed to overlap, and >95% of patients diagnosed with NAFLD in Japan are also diagnosed with MASLD [23, 24]. In this study, of the 14,141 individuals who met the diagnostic criteria for NAFLD in the cross-sectional data, 12,632 (89.33%) were also diagnosed with MASLD, but the overlap rate was lower than previous reports. A possible reason is that our study participants were relatively young and had a lower BMI. In a group with a larger proportion of individuals with normal BMI, the likelihood of not meeting the diagnostic criteria for MASLD (BMI ≥23) becomes higher, which may contribute to an increase in the proportion of patients diagnosed with NAFLD but not with MASLD. Furthermore, patients who were not obese and had SLD were more strongly influenced by the PNPLA3 genotype, which is thought to contribute to SLD development [25]. In patients who were not obese, the degree of metabolic disturbance is often lower, which indicates that the proportion of patients diagnosed with NAFLD but not with MASLD may be higher in individuals who were not obese. In fact, approximately 90% of patients who were not obese and had NAFLD were diagnosed with MASLD, which is comparable to the proportion observed in this study. Thus, the overlap rate between NAFLD and MASLD may be influenced by the degree of obesity in the population, and further research is required to investigate the overlap rate by different levels of obesity.
High UA levels have been reported to be significantly associated with NAFLD development [26-30]. Given many patients with NAFLD and MASLD overlap, UA levels may also serve as a diagnostic predictive marker for MASLD. Indeed, He et al. demonstrated that UA levels could be a predictive factor for MASLD development in their analysis of 16,152 Chinese individuals without steatotic liver, with a BMI <25, and with low alcohol consumption (<140 g/week for men and <70 g/week for women) [31]. Similar to He et al., we reported that UA levels were a predictive factor for MASLD and MetALD development. Conversely, the present study demonstrated that, unlike the study by He et al., UA levels were not an independent predictive marker for MASLD/MetALD development in participants with a BMI >25, regardless of sex. This indicates that UA may not serve as a predictive marker for MASLD/MetALD development depending on patient characteristics. Furthermore, UA similarly served as a predictive factor for MetALD development in men who consume a moderate amount of alcohol (210–420 g/week). Considering that even low alcohol consumption contributes to SLD development and progression, and that alcohol consumption increases the risk of hyperuricemia in men, alcohol intake will highly likely act as a confounding factor in the relationship between UA levels and MASLD/MetALD development in men [32, 33]. However, even after adjusting for alcohol consumption in men, UA levels maintained an independent association with the risk of MASLD/MetALD development. On the contrary, in women, UA levels did not predict MASLD/MetALD development in those with a certain level of alcohol consumption (140–350 g/week). This may be attributed to the small proportion of women with such alcohol consumption (1,431 out of 11,913) and the low number of MASLD/MetALD cases during follow-up (n = 182), which may have reduced the power to detect differences. Therefore, further large-scale studies are needed to determine whether UA levels could be a predictive marker for MASLD/MetALD development in women with certain levels of alcohol consumption.
Similar to previous studies evaluating the association of UA levels with NAFLD, the present study demonstrated that UA is an independent predictive factor for MASLD/MetALD development. On the contrary, whether hyperuricemia itself contributes to the development or progression of both insulin resistance and SLD remains unclear. Several mechanisms by which hyperuricemia can cause SLD were hypothesized. Endoplasmic reticulum stress induced by hyperuricemia was reported to activate SREBP-1, leading to liver fat accumulation [34]. Another study revealed that UA induced insulin resistance and impaired insulin signaling through a reactive oxygen species-related pathway, both in vivo and in vitro [35]. Furthermore, a mouse study demonstrated that UA contributes to insulin resistance and hepatocyte fat accumulation through the activation of the NOD-like receptor family pyrin domain containing 3 inflammasome [36]. Furthermore, UA suppresses AMP kinase activity in the hepatocytes and decreases fatty acid oxidation, accelerating fat accumulation in the liver [37]. In a study of diet-induced obese mice, Tanaka et al. reported that the use of a URAT1-selective inhibitor to lower UA levels in mice resulted in an improvement in SLD [38]. These findings indicate that UA may contribute to SLD worsening; however, in humans, evidence supporting the role of hyperuricemia in SLD onset or progression is highly limited. A large-scale study using Mendelian randomization did not provide evidence that high UA levels contributed to NAFLD development or progression [39]. Whether these findings apply similarly to MASLD and MetALD remains unclear at this time; therefore, further studies are needed.
This study has several limitations. First, data were obtained from health checkups at three clinics, which may restrict the broader applicability of our findings. In addition, because the participants were exclusively Japanese, the relevance of these results to other ethnic groups remains unverified. Second, MASLD/MetALD was diagnosed using ultrasonography, which can underestimate hepatic steatosis prevalence when it is <20% [40]. Performing biopsies, which are the gold standard for diagnosis, is also impractical for all patients suspected of SLD undergoing routine health checkups. Third, given the retrospective nature of this study, selection bias may have occurred. It remains to be determined through large-scale prospective studies whether UA is a causative factor for the development of MASLD/MetALD or merely serves as a clinical marker to identify patients at high risk of developing these conditions. Finally, many individuals tend to underestimate their alcohol consumption when self-reporting using questionnaires [41]. Therefore, the relationship between alcohol consumption and MASLD and between alcohol consumption and MetALD development may not be accurately assessed. Further studies are needed to investigate the association between alcohol consumption and MASLD and MetALD development.
Conclusion
As summarized in the Graphical Abstract, this study highlights a significant association between high UA levels and MASLD/MetALD development. In addition, high UA levels are associated with MASLD/MetALD development after adjustment for other CMRFs, proposing that high UA levels are a sensitive predictive marker for future MASLD/MetALD development. This underscores that UA levels measured during health checkups may be useful for predicting the risk of MASLD/MetALD development and early identification and management of these conditions.
Graphical Abstract
Author Contributions
T.Fukuda: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. T.A: Investigation, Writing – original draft, Writing – review & editing. T.O: Supervision, Validation. T.Fukaishi: Supervision, Validation. A.K: Supervision, Validation. M.T: Supervision, Validation. T.Y: Funding acquisition, Project administration, Software. K.M: Project administration, Supervision, Visualization.
Funding
This study was supported by fund of Department of Molecular Endocrinology and Metabolism, Graduate School of Medical and Dental Sciences, Institute of Science Tokyo.
Acknowledgments
The authors thank Enago (www.enago.jp) for the English language review.
Disclosure
None of the authors have any potential conflicts of interest associated with this research.
Data Availability Statement
Data generated or analyzed during the study are available from the corresponding author upon reasonable request.
Supplementary Fig. 1
ROC curves for predicting MASLD or MetALD in all participants, men, and women stratified by alcohol consumption. MASLD, metabolic dysfunction-associated steatotic liver disease; MetALD, metabolic and alcohol related/associated liver disease.
Supplementary Fig. 2
Kaplan–Meier curve of the time to the development of MASLD/MetALD according to the quartiles of UA levels for each of CMRF 1 (A), 2 (B), 3 (C), and 4 or more (D) in men.
Supplementary Fig. 3
Kaplan–Meier curve of the time to the development of MASLD/MetALD according to the quartiles of UA levels for each of CMRF 1 (A), 2 (B), 3 (C), and 4 or more (D) in women.
References
- 1 Younossi ZM, Koenig AB, Abdelatif D, Fazel Y, Henry L, et al. (2016) Global epidemiology of nonalcoholic fatty liver disease-Meta-analytic assessment of prevalence, incidence, and outcomes. Hepatology 64: 73–84.
- 2 Wang XJ, Malhi H (2018) Nonalcoholic fatty liver disease. Ann Intern Med 169: ITC65–ITC80.
- 3 Tomic D, Kemp WW, Roberts SK (2018) Nonalcoholic fatty liver disease: current concepts, epidemiology and management strategies. Eur J Gastroenterol Hepatol 30: 1103–1115.
- 4 Younossi ZM, Golabi P, Paik JM, Henry A, Van Dongen C, et al. (2023) The global epidemiology of nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH): a systematic review. Hepatology 77: 1335–1347.
- 5 Younossi ZM (2019) Non-alcoholic fatty liver disease—A global public health perspective. J Hepatol 70: 531–544.
- 6 Ge X, Zheng L, Wang M, Du Y, Jiang J (2020) Prevalence trends in non-alcoholic fatty liver disease at the global, regional and national levels, 1990–2017: a population-based observational study. BMJ Open 10: e036663.
- 7 Idalsoaga F, Kulkarni AV, Mousa OY, Arrese M, Arab JP (2020) Non-alcoholic fatty liver disease and alcohol-related liver disease: two intertwined entities. Front Med (Lausanne) 7: 448.
- 8 Rinella ME, Lazarus JV, Ratziu V, Francque SM, Sanyal AJ, et al. (2023) A multisociety Delphi consensus statement on new fatty liver disease nomenclature. Hepatology 78: 1966–1986.
- 9 Valenti L, Bugianesi E, Pajvani U, Targher G (2016) Nonalcoholic fatty liver disease: cause or consequence of type 2 diabetes? Liver Int 36: 1563–1579.
- 10 Muzurović E, Mikhailidis DP, Mantzoros C (2021) Non-alcoholic fatty liver disease, insulin resistance, metabolic syndrome and their association with vascular risk. Metabolism 119: 154770.
- 11 Bombelli M, Quarti-Trevano F, Tadic M, Facchetti R, Cuspidi C, et al. (2018) Uric acid and risk of new-onset metabolic syndrome, impaired fasting glucose and diabetes mellitus in a general Italian population: data from the Pressioni Arteriose Monitorate E Loro Associazioni study. J Hypertens 36: 1492–1498.
- 12 Li C, Hsieh MC, Chang SJ (2013) Metabolic syndrome, diabetes, and hyperuricemia. Curr Opin Rheumatol 25: 210–216.
- 13 Ali N, Miah R, Hasan M, Barman Z, Mou AD, et al. (2020) Association between serum uric acid and metabolic syndrome: a cross-sectional study in Bangladeshi adults. Sci Rep 10: 7841.
- 14 Krishnan E, Pandya BJ, Chung L, Hariri A, Dabbous O (2012) Hyperuricemia in young adults and risk of insulin resistance, prediabetes, and diabetes: a 15-year follow-up study. Am J Epidemiol 176: 108–116.
- 15 Utzschneider KM, Kahn SE (2006) Review: the role of insulin resistance in nonalcoholic fatty liver disease. J Clin Endocrinol Metab 91: 4753–4761.
- 16 Sun Q, Zhang T, Manji L, Liu Y, Chang Q, et al. (2023) Association between serum uric acid and non-alcoholic fatty liver disease: an updated systematic review and meta-analysis. Clin Epidemiol 15: 683–693.
- 17 Sterling RK, Lissen E, Clumeck N, Sola R, Correa MC, et al. (2006) Development of a simple noninvasive index to predict significant fibrosis in patients with HIV/HCV coinfection. Hepatology 43: 1317–1325.
- 18 Fukuda T, Okamoto T, Fukaishi T, Kawakami A, Tanaka M, et al. (2024) Extent to which weight loss contributes to improving metabolic dysfunction-associated and metabolic and alcohol related/associated steatotic liver disease: a study on Japanese participants undergoing health checkups. Front Endocrinol (Lausanne) 15: 1392280.
- 19 Bedogni G, Bellentani S, Miglioli L, Masutti F, Passalacqua M, et al. (2006) The fatty liver index: a simple and accurate predictor of hepatic steatosis in the general population. BMC Gastroenterol 6: 33.
- 20 The Examination Committee of Criteria for ‘Obesity Disease’ in Japan; Japan Society for the Study of Obesity (2002) New criteria for ‘obesity disease’ in Japan. Circ J 66: 987–992.
- 21 Fluss R, Faraggi D, Reiser B (2005) Estimation of the Youden index and its associated cutoff point. Biom J 47: 458–472.
- 22 Hak AE, Choi HK (2008) Menopause, postmenopausal hormone use and serum uric acid levels in US women—the third national health and nutrition examination survey. Arthritis Res Ther 10: R116.
- 23 Kamada Y, Fujii H, Suzuki Y, Sawada K, Tatsuta M, et al. (2024) Clinical characteristics of steatotic liver disease categories in a large cohort of Japanese health checkup participants. Gastro Hep Adv 3: 1148–1156.
- 24 Suzuki K, Tamaki N, Kurosaki M, Takahashi Y, Yamazaki Y, et al. (2024) Concordance between metabolic dysfunction-associated steatotic liver disease and nonalcoholic fatty liver disease. Hepatol Res 54: 600–605.
- 25 Seko Y, Lin H, Wong VW, Okanoue T (2024) Impact of PNPLA3 in Lean Individuals and in Cryptogenic Steatotic Liver Disease. Liver Int doi: 10.1111/liv.16164. Online ahead of print.
- 26 Xu C, Yu C, Xu L, Miao M, Li Y (2010) High serum uric acid increases the risk for nonalcoholic Fatty liver disease: a prospective observational study. PLoS One 5: e11578.
- 27 Lee JW, Cho YK, Ryan M, Kim H, Lee SW, et al. (2010) Serum uric acid as a predictor for the development of nonalcoholic fatty liver disease in apparently healthy subjects: a 5-year retrospective cohort study. Gut Liver 4: 378–383.
- 28 Wei F, Li J, Chen C, Zhang K, Cao L, et al. (2020) Higher serum uric acid level predicts non-alcoholic fatty liver disease: a 4-year prospective cohort study. Front Endocrinol (Lausanne) 11: 179.
- 29 Darmawan G, Hamijoyo L, Hasan I (2017) Association between serum uric acid and non-alcoholic fatty liver disease: a meta-analysis. Acta Med Indones 49: 136–147.
- 30 Wijarnpreecha K, Panjawatanan P, Lekuthai N, Thongprayoon C, Cheungpasitporn W, et al. (2017) Hyperuricaemia and risk of nonalcoholic fatty liver disease: a meta-analysis. Liver Int 37: 906–918.
- 31 He L, Qiu K, Zheng W, Kong W, Zeng T (2024) Uric acid may serve as the sixth cardiometabolic criterion for defining MASLD. J Hepatol 80: e152–e153.
- 32 Long MT (2022) Alcohol consumption and nonalcoholic steatohepatitis/nonalcoholic fatty liver disease. Gastroenterol Hepatol (N Y) 18: 491–493.
- 33 Li Z, Guo X, Liu Y, Chang Y, Sun Y, et al. (2016) The relation of moderate alcohol consumption to hyperuricemia in a rural general population. Int J Environ Res Public Health 13: 732.
- 34 Choi YJ, Shin HS, Choi HS, Park JW, Jo I, et al. (2014) Uric acid induces fat accumulation via generation of endoplasmic reticulum stress and SREBP-1c activation in hepatocytes. Lab Invest 94: 1114–1125.
- 35 Zhi L, Yuzhang Z, Tianliang H, Hisatome I, Yamamoto T, et al. (2016) High uric acid induces insulin resistance in cardiomyocytes in vitro and in vivo. PLoS One 11: e0147737.
- 36 Wan X, Xu C, Lin Y, Lu C, Li D, et al. (2016) Uric acid regulates hepatic steatosis and insulin resistance through the NLRP3 inflammasome-dependent mechanism. J Hepatol 64: 925–932.
- 37 Lanaspa MA, Cicerchi C, Garcia G, Li N, Roncal-Jimenez CA, et al. (2012) Counteracting roles of AMP deaminase and AMP kinase in the development of fatty liver. PLoS One 7: e48801.
- 38 Tanaka Y, Nagoshi T, Takahashi H, Oi Y, Yoshii A, et al. (2022) URAT1-selective inhibition ameliorates insulin resistance by attenuating diet-induced hepatic steatosis and brown adipose tissue whitening in mice. Mol Metab 55: 101411.
- 39 Li S, Fu Y, Liu Y, Zhang X, Li H, et al. (2022) Serum uric acid levels and nonalcoholic fatty liver disease: a 2-sample bidirectional mendelian randomization study. J Clin Endocrinol Metab 107: e3497–e3503.
- 40 Khov N, Sharma A, Riley TR (2014) Bedside ultrasound in the diagnosis of nonalcoholic fatty liver disease. World J Gastroenterol 20: 6821–6825.
- 41 Livingston M, Callinan S (2015) Underreporting in alcohol surveys: whose drinking is underestimated? J Stud Alcohol Drugs 76: 158–164.
