Prevalence and risk factors for metabolic dysfunction-associated steatotic liver disease and fibrosis in recently diagnosed type 2 diabetes

Abstract

BACKGROUND

Metabolic dysfunction-associated steatotic liver disease (MASLD) is the most common chronic liver disease and the leading cause of liver-related mortality. Long-standing type 2 diabetes mellitus (T2DM) is a major risk factor for liver steatosis and fibrosis; however, whether recently diagnosed T2DM carries the same risk remains unclear. Few studies have addressed this question in Hispanic populations, who are known to exhibit faster progression of liver disease.

AIM

To assess the prevalence of MASLD and liver fibrosis in patients with recent-onset T2DM and to identify associated risk factors.

METHODS

This was a cross-sectional study including individuals with T2DM diagnosed within the past 5 years and without disabling micro or macrovascular complications. Clinical, biochemical, liver stiffness measurement (LSM), controlled attenuation parameter (CAP), and FibroScan-AST (FAST) score were evaluated.

RESULTS

One thousand one hundred and thirty-seven patients were included, 60.9% were women, median age was 55 years (inter quartile range: 49-63) and 78.2% were overweight or obese. Overall prevalence of MASLD (using the > 288 dB/m cutoff) was 41.4%, and 10% had clinically significant liver fibrosis (LSM ≥ 8 kPa). Moreover, 3% of the population had cirrhosis. Multivariate analysis showed higher body mass index (BMI), glycated hemoglobin, triglycerides and alanine aminotransferase (ALT) were independently associated with steatosis, while higher BMI, CAP, gamma-glutamyl transferase, ALT and aspartate aminotransferase were associated with fibrosis. According to the FAST score, 4.8% of patients had at-risk metabolic dysfunction-associated steatohepatitis.

CONCLUSION

There is a high burden of MASLD and clinically significant fibrosis among adults with recently diagnosed T2DM. These findings strongly support the systematic evaluation for MASLD and liver fibrosis in this population from the time of diabetes diagnosis.

Key Words: Metabolic dysfunction-associated steatotic liver disease; Type 2 diabetes mellitus; Steatotic liver disease; Liver fibrosis

Core Tip: There is scarce data specifically evaluating the prevalence of metabolic dysfunction-associated steatotic liver disease (MASLD) and liver fibrosis in patients with recent-onset type 2 diabetes (< 5 years since diagnosis) and without micro or macrovascular complications. The Hispanic population is regarded as high-risk for metabolic diseases, including MASLD. We found a high prevalence of MASLD, affecting 41.4% of the study population, with 10% presenting clinically significant liver fibrosis (liver stiffness measurement ≥ 8 kPa) and 3% having cirrhosis. Our findings support incorporating liver health screening into the routine evaluation of patients with type 2 diabetes mellitus from diagnosis, even in those with good glycemic control and without chronic complications.

INTRODUCTION

Metabolic dysfunction-associated steatotic liver disease (MASLD) is the most common cause of chronic liver disease and one of the leading causes of liver-related mortality, with an estimated overall prevalence of 30% in the adult population[1,2]. A 2022 meta-analysis reported an increase in global prevalence from 25.5% in 2005 to 37.8% in 2016, with an estimated incidence of 46.9 cases per 1000 person/years[3]. MASLD prevalence appears to differ across ethnic groups; in United States studies, Hispanic individuals exhibit higher rates (around 23%) than White (14%) and Black (13%) populations; however, specific data on Hispanic populations outside the United States is still limited[1].

MASLD is strongly associated with metabolic disorders, including dyslipidemia, central obesity, hypertension, obstructive sleep apnea-hypopnea syndrome, chronic kidney disease, cardiovascular disease, insulin resistance, and especially diabetes. Diabetes is a major risk factor for both liver steatosis and fibrosis. Notably, metabolic dysfunction-associated steatohepatitis (MASH) prevalence in this population exceeds 33%, with approximately 17% of patients presenting advanced liver fibrosis[4,5].

Liver fibrosis in MASLD is the strongest predictor of adverse clinical outcomes, including liver-related and extrahepatic mortality. The risk of liver-related death rises with each fibrosis stage (F1 to F4) and is the primary determinant of outcomes in MASLD patients[6-9].

Although the risk of liver steatosis and fibrosis in diabetes is well-documented, data on the Mexican population remains limited. This is particularly relevant given the high prevalence of risk polymorphisms in Hispanics. Moreover, most studies have not assessed MASLD prevalence in newly diagnosed type 2 diabetes mellitus (T2DM) patients. This is crucial, as the likelihood of advanced fibrosis increases with diabetes duration. Early-stage T2DM presents an opportunity for intensive treatment, which can yield long-term benefits by modifying liver outcomes, slowing disease progression, and ultimately improving clinical outcomes by reducing the risk of cirrhosis, its complications, and both liver-related and non-liver-related mortality[10].

Therefore, the aim of this study was to determine the prevalence and risk factors associated with liver steatosis, fibrosis, and at-risk MASH using controlled attenuation parameter (CAP) and vibration-controlled transient elastography (VCTE) in patients with recently diagnosed T2DM.

MATERIALS AND METHODS

Study design

This is a cross-sectional study carried out at a tertiary hospital in Mexico City from January 2022 to January 2024. The study followed the principles outlined in the Declaration of Helsinki and was approved by the Research Ethics Committee of the institution (No. 2985).

Selection of patients

Patients older than 18 years with a recent diagnosis (less than 5 years) of T2DM, without disabling complications (blindness, renal failure, stroke, limb amputations, ischemic heart disease), who attended our specialized diabetes care clinic, were included. Patients with a history of liver diseases, including viral hepatitis, autoimmune liver diseases, and alcohol-related liver disease, were excluded from the analysis.

Specialized diabetes care clinic-center of comprehensive care for the patient with diabetes

The center of comprehensive care for the patient with diabetes clinic is a program aimed at improving T2DM management and preventing complications. It serves patients aged 18 to 70, diagnosed within the past 5 years. The clinic offers a multidisciplinary approach, providing self-care tools, preventing complications, and supporting long-term diabetes management. Patients receive care from endocrinologists, diabetes educators, nutritionists, psychologists, psychiatrists, dental specialists, physical activity instructors, ophthalmologists, and foot care specialists. Treatment is tailored by the attending physician without restrictions on antidiabetic drugs.

During the participants’ assessment, psychologists conducted a structured interview to evaluate the quantity and frequency of alcohol consumption. Moderate alcohol consumption was defined as 140 g/week in women and 210 g/week in men.

VCTE

All patients included in the study underwent VCTE performed by one of three highly experienced physicians (each with more than 1500 studies). Evaluations were conducted using a FibroScan^® Expert 630 device (Echosens, Paris, France) and met the quality criteria required for valid liver stiffness measurement (LSM) and CAP measurements, with interquartile range (IQR) < 20%[11,12].

To evaluate the presence of steatosis, we used a CAP cutoff of ≥ 288 dB/m, as recommended by the American Association for the Study of Liver Diseases (AASLD) guidelines[12]. We further categorized the patients according to the risk for liver fibrosis as low (< 8 kPa), intermediate (8-12 kPa), and high-risk (> 12 kPa)[2]. Accordingly, we defined clinically significant liver fibrosis as a LSM ≥ 8 kPa. Specifically, a LSM ≥ 13.6 kPa cutoff was used to define F4 (LSM compatible with cirrhosis).

In addition to VCTE, the FibroScan-AST (FAST) score was calculated as previously reported[13]. This score incorporates values of LSM, CAP, and serum aspartate aminotransferase (AST) concentrations, with the aim of non-invasively identifying patients with suspected MASH who are at risk of progression. Two cut-off points were used in prior studies: ≤ 0.35 to rule out its presence (with 90% sensitivity) and ≥ 0.67 to rule it in (with 90% specificity)[13,14]. An adjusted cut-off point of ≥ 0.5 was also included for patients with glycated hemoglobin (HbA1c) ≥ 6.5% in the analysis, due to the higher proportion of patients at-risk MASH when HbA1c ≥ 6.5%[15]. The rationale for this adjustment is that patients with T2DM have increased oxidative and endoplasmic reticulum stress, intestinal dysbiosis, and adipose tissue dysfunction, promoting liver steatosis and fibrosis[16].

In the present study, information from studies under the definition of NAFLD was used, corresponding to the recent consensus definition of MASLD.

Biochemical variables

Laboratory tests were performed after an 8-hour fasting period. Samples were immediately centrifuged, and blood chemistry, alanine aminotransferase (ALT), AST, gamma-glutamyl transferase (GGT), lipid profile, and HbA1c levels were determined on the same day, following the quality standards set by our central laboratory, accredited by the College of American Pathologists.

Statistical analysis

The sample size was calculated based on a prevalence of liver steatosis of 35% in patients without diabetes and 40% in patients with diabetes. Using alpha and beta errors of 5% and 80%, respectively, the required sample size was 742 patients. After accounting for a 10% loss, the final sample size was adjusted to 817 patients. However, a total of 1337 patients were included in this study.

Kolmogorov-Smirnov test was used to evaluate data distribution. To describe the characteristics of the population, means ± SD and medians (IQR) were used for quantitative data, while proportions and relative frequencies were used for categorical variables. To compare groups based on the presence of steatosis and liver fibrosis, either the student t-test or Mann-Whitney U test was used, according to the distribution of the data. For categorical variables, the χ² test or Fisher’s exact test was used as appropriate. Logistic regression analysis was performed to establish the factors independently associated with the presence of steatosis and liver fibrosis. Variables were selected a priori, based on clinical and biological plausibility. To evaluate the validity of the multivariable models (Supplementary Table 1) collinearity among covariables was assessed using variance inflation factors, model performance was evaluated using the number of events per variable, discrimination using χ² statistic with 95% confidence interval, and calibration slope. To obtain estimates corrected for optimism, internal validation was performed using 1000 bootstrap resamples. A P value < 0.05 was considered statistically significant. Statistical analysis was performed using SPSS version 25 (IBM, Armonk NY, United States) and STATA version 12.1 (Stata Corp LLC, College Station, TX, United States).

RESULTS

General characteristics of the study population

A total of 1337 patients with a recent diagnosis of T2DM were included in the analysis. The majority of the patients were women (n = 814, 60.9%), with a median age of 55 years, and only 21.8% of the population had normal body mass index (BMI) (< 25 kg/m²).

Alcohol consumption within the population

Regarding alcohol consumption, 74.3% of the study population denied any alcohol intake. Only six patients (0.5%) had a history of drinking more than three days per week. The median (p25-p75) alcohol intake in the population was 0 (0-7.2) g/week. A total of 50 patients (3.9%) reported consuming alcohol at least once per week, with a median (p25-p75) alcohol consumption of 28.6 (15.9-48.4) g/week. Nine patients (7 men and two women) had moderate alcohol consumption, corresponding to 0.6% of the population, and were classified as MASLD and increased alcohol intake. Therefore, most of the cohort corresponded to MASLD.

Prevalence and risk factors associated with the presence of liver steatosis

The prevalence of MASLD was 41.4%, (554 patients), Table 1 presents the characteristics of the study population based on the absence or presence of MASLD. Patients with steatosis were younger and had a higher BMI; notably, they also had higher LSM and FAST scores.

Table 1 Characteristics of the study population according to presence or absence of liver steatosis, n (%).

	Total (n = 1337)	No steatosis (n = 783)	Steatosis (n = 554)	P value
Sex (female)	814 (60.9)	468 (59.8)	346 (62.5)	0.322
Age (years)	55 (49-63)	56 (49-63)	54 (48-61)	0.003
Classification according to BMI
BMI (kg/m2), median (p25-p75)	28.3 (25.4-31.5)	27.1 (24.5-30.1)	30.4 (27.2-33.3)	< 0.001
Normal	276 (21.8)	217 (29.3)	59 (11.2)	< 0.001
Overweight	510 (40.2)	325 (43.9)	185 (35.0)
Obesity class I	346 (27.3)	153 (20.7)	193 (36.6)
Obesity class II	99 (7.8)	35 (4.7)	64 (12.1)
Obesity class III	37 (2.9)	10 (1.4)	27 (5.1)
Transient elastography
kPa	4.8 (3.8-6.0)	4.5 (3.5-5.6)	5.2 (4.2-6.6)	< 0.001
CAP (dB/m)	276 (234-317)	242 (214-266)	324 (305-346)	< 0.001
FAST score	0.07 (0.03-0.18)	0.05 (0.02-0.10)	0.15 (0.06-0.29)	< 0.001
Low risk	984 (88.0)	628 (94.3)	356 (78.8)	< 0.001
Moderate risk	109 (9.7)	32 (4.8)	77 (17.0)
High risk	25 (2.2)	6 (0.9)	19 (4.2)
FAST-HbA1c	54 (4.8)	11 (1.7)	43 (9.5)	< 0.001
Biochemical parameters
Total cholesterol, mg/dL	165 (137-193)	163 (134-190)	168 (140-197)	0.036
LDL, mg/dL	98 (74-120)	96 (73-117)	100 (77-124)	0.022
HDL, mg/dL	42 (36-51)	43 (37-52)	40 (36-48)	0.000
TGC, mg/dL	139(104-194)	127 (97-177)	157 (117-224)	0.000
Creatinine, mg/dL	0.76(0.65-0.90)	0.77 (0.67-0.90)	0.73 (0.63-0.89)	0.010
ALT, U/L	23.5(17.0-34.0)	20.7 (15.7-30.3)	28.6 (19.4-41.9)	0.000
AST, U/L	20.7 (16.6-26.9)	19.4 (16.0-25.0)	22.3 (17.7-29.2)	0.000
GGT, U/L	25.9(18.4-40.4)	22.7 (17.2-33.5)	31.8 (22.2-49.2)	0.000
HbA1c, %	6.8 (6.0-8.5)	6.6 (6.0-8.4)	7.0 (6.1-8.7)	0.018

The results are expressed as median (p25-p75). Low risk (< 0.35), moderate risk (0.35-0.66), high risk ( 0.67). BMI: Body mass index; FAST: FibroScan-AST; LDL: Low-density cholesterol; HDL: High-density cholesterol; TGC: Triglycerides; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; GGT: Gamma-glutamyl transferase; CAP: Controlled attenuation parameter; HbA1c: Glycated hemoglobin.

Patients with liver steatosis also exhibited significant differences in blood chemistry values, including higher levels of triglycerides, ALT, AST, GGT, and HbA1c, as well as lower serum high-density cholesterol (HDL) levels and higher total and low-density cholesterol (LDL) cholesterol levels.

Prevalence and risk factors associated with the presence of clinically significant liver fibrosis

Table 2 shows the characteristics of the study population based on the fibrosis risk assessment. The prevalence of fibrosis was 10%, with 133 patients having LSM 8 kPa indicating clinically significant liver fibrosis ( F2). Remarkably, 40 patients (3%) met the criteria for cirrhosis (F4). None of these patients had a prior diagnosis of liver disease before this assessment.

Table 2 Characteristics of the study population according to the risk of liver fibrosis, n (%).

	< 8 kPa (n = 1204)	8-12 kPa (n = 77)	> 12 kPa (n = 56)	P value
Sex (female)	729 (60.5)	52 (67.5)	33 (58.9)	0.455
Age (years)	55 (49-63)	56 (48.5-62)	58 (52-64)	0.167
BMI (kg/m2)	27.9 (25.2-31.0)	31.8 (28.2-36.2)	31.4 (27.8-35.0)	< 0.001a,b
BMI classification
Normal	274 (22.8)	9 (11.7)	8 (14.3)	< 0.001
Overweight	510 (42.4)	20 (26)	14 (25)
Obesity	317 (26.3)	25 (32.5)	20 (35.7)
Obesity class II	78 (6.5)	16 (20.8)	8 (14.3)
Obesity class III	25 (2.1)	7 (9.1)	6 (10.7)
Transient elastography
kPa	4.6 (3.7-5.6)	9 (8.5-10.3)	16.5 (13.4-21.8)	< 0.001a,b,c
CAP (dB/m)	273 (231-314)	303 (266-330)	298 (257 – 349)	< 0.001a,b
FAST score1	0.06 (0.03-0.14)	0.41 (0.21-0.59)	0.56 (0.41-0.77)	< 0.001a,b,c
Low risk	947 (94)	31 (47.7)	(13)
Moderate risk	59 (5.9)	29 (44.6)	(45.7)
High risk	1 (0.1)	5 (7.7)	19 (41.3)
FAST-HbA1c	8 (0.7)	20 (26)	26 (46.4)	< 0.001
Biochemical parameters
Total cholesterol, mg/dL	166 (138-195)	161 (131-180)	152 (128-177)	0.028b
LDL, mg/dL	98 (75-121)	101 (68-116)	91 (68-104)	0.124
HDL, mg/dL	42 (36-51)	39 (35-47)	42 (35-48)	0.085
TGC, mg/dL	140 (103-195)	122 (102-189)	135 (114-181)	0.666
Creatinine mg/dL	0.77 (0.65-0.90)	0.73 (0.62-0.84)	0.73 (0.59-0.88)	0.107
ALT, U/L	22.4 (16.5-32.4)	39.9 (26.3-65.3)	40.2 (30.4-74.1)	< 0.001a,b
AST, U/L	19.7 (16.2-25.2)	30.5 (22.2-51.4)	33.5 (28.3-57.3)	< 0.001a,b
GGT, U/L	24.8 (17.9-35.8)	42.3 (28.9-64.6)	85.9 (51.1-144.8)	< 0.001a,b,c
HbA1c, %	6.7 (6-8.5)	7.1 (6.3 – 8.9)	7.2 (6.2-9.4)	0.074

¹Available in 1118 patients.

^aP < 0.05 vs group 1 and 2.

^bP < 0.05 vs group 1 and 3.

^cP < 0.05 vs group 2 and 3.

Group 1 (< 8 kPa), group 2 (8-12 kPa) and group 3 (> 12 kPa). The results are expressed as median (p25-p75). BMI: Body mass index; FAST: FibroScan-AST; LDL: Low density cholesterol; HDL: High density cholesterol; TGC: Triglycerides; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; GGT: Gamma-glutamyl transferase; CAP: Controlled attenuation parameter; HbA1c: Glycated hemoglobin.

Patients with clinically significant fibrosis had higher BMI and CAP values than those without fibrosis. Additionally, they had higher ALT, AST, and GGT concentrations and lower total cholesterol levels. LDL cholesterol, HDL cholesterol, triglycerides, and hemoglobin A1c levels showed no significant differences.

Lean MASLD

Supplementary Table 2 shows the characteristics of patients with lean MASLD (10.6% of patients with MASLD) compared with MASLD, apart from the expected difference in BMI, there were no significant differences. Female gender was predominantly seen in both categories, kPa and FAST score were similar, and there was a non-significant trend towards higher CAP in MASLD compared with lean MASLD, but this seems as a non-clinically significant trend.

Apart from that, there were only trends towards significance in age and total cholesterol, both being higher in the lean MASLD group.

Non-invasive evaluation according to HbA1C levels

Table 3 presents the characteristics of the study population based on glycemic control categories. The population was categorized into adequate (HbA1c ≤ 7%) and inadequate (HbA1c > 7%) glycemic control groups. Most patients (53.3%) had adequate glycemic control. Patients with inadequate glycemic control had higher liver stiffness and CAP values. As expected, they also had higher cholesterol and triglyceride levels, lower HDL, and elevated GGT levels.

Table 3 Characteristics of the study population according to control of type 2 diabetes mellitus according to glycated hemoglobin values, n (%).

	≤ 7% HbA1c (n = 649)	> 7% HbA1c (n = 508)	P value
Sex (female)	420 (64.7)	271 (53.3)	< 0.001
Age (years)	56 (49-63)	54 (47-62)	0.060
BMI (kg/m2)	28.1 (25.3-31.5)	28.4 (25.4-31.5)	0.621
BMI classification
Normal	135 (21.8)	102 (21.3)	0.496
Overweight	252 (40.8)	192 (40.0)
Obesity class I	169 (27.3)	130 (27.1)
Obesity class II	51 (8.3)	39 (8.1)
Obesity class III	11 (1.8)	17 (3.5)
Transient elastography
kPa	4.6 (3.7-5.9)	4.9 (3.9-6.2)	0.015
CAP (dB/m)	271 (231-314)	282 (237-320)	0.040
FAST score	0.08 (0.03-0.17)	0.07(0.03-0.19)	0.530
Low risk	530 (87.9)	430 (88.3)	0.077
Moderate risk	64 (10.6)	41 (8.4)
High risk	9 (1.5)	16 (3.3)
Biochemical parameters
Total cholesterol, mg/dL	163 (137-186)	171 (138-200)	0.002
LDL, mg/dL	97 (75-117)	99 (75-125)	0.086
HDL, mg/dL	43 (37-51)	40 (35-49)	< 0.001
TGC, mg/dL	129 (98-175)	159 (115-233)	< 0.001
Creatinine, mg/dL	0.76 (0.66-0.90)	0.76 (0.64-0.89)	0.297
ALT, U/L	22.7 (16.7-33.5)	24.1 (17.2-34.8)	0.091
AST, U/L	20.9 (17.2-27.3)	20.0 (15.7-26.6)	0.027
GGT, U/L	23.8 (17.5-35.6)	28.9 (20.4-45.5)	< 0.001
HbA1c, %	6.1(5.7-6.5)	8.9 (7.7-10.5)	< 0.001

The results are expressed as median (p25-p75). 1157 patients had HbA1c determined at the time of elastography. BMI: Body mass index; FAST: FibroScan-AST; LDL: Low density cholesterol; HDL: High density cholesterol; TGC: Triglycerides; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; GGT: Gamma-glutamyl transferase; CAP: Controlled attenuation parameter; HbA1c: Glycated hemoglobin.

To further investigate the impact of glycemic control on TE-related parameters, an analysis using different HbA1c cutoffs is presented in Supplementary Table 3. Notable findings include a higher proportion of male patients among those with elevated HbA1c levels. Additionally, liver stiffness showed a gradual, though insignificant, increase, while CAP levels and high-risk FAST scores were higher in these groups.

In terms of liver chemistry, GGT levels exhibited a stepwise increase as HbA1c levels rose. Although differences in transaminase levels were observed across groups, no specific trend associated with HbA1c levels was identified.

Multivariate analysis

Independent risk factors for the presence of liver steatosis: Table 4 presents the results of a logistic regression analysis evaluating risk factors independently associated with liver steatosis. In the Univariate analysis, age, BMI class, HbA1c ≥ 6.5%, triglycerides, LDL-C, and abnormal transaminases were associated with the presence of liver steatosis.

Table 4 Logistic regression to evaluate the risk factors associated with presence of liver steatosis.

Univariate	OR	95%CI	β	P value
Age (years)	0.988	0.977-0.999	-0.12	0.027
BMI class	1.890	1.665-2.146	0.637	< 0.001
Sex	0.960	0.763-1.208	-0.041	0.727
HbA1c ≥ 6.5%	1.547	1.205-1.986	0.436	< 0.001
Triglycerides (mg/dL)	1.003	1.002-1.005	0.003	< 0.001
LDL cholesterol (mg/dL)	1.004	1.000-1.007	0.004	0.034
High AST	2.304	1.542-3.442	0.835	< 0.001
High ALT	2.634	2.044-3.395	0.969	< 0.001
Multivariate
BMI class	1.763	1.522-2.042	0.567	< 0.001
Triglycerides, mg/dL	1.002	1.001-1.004	0.002	< 0.001
LDL cholesterol	1.004	1.000-1.008	0.004	0.065
HbA1c ≥ 6.5%	1.541	1.153-2.060	0.432	0.004
High ALT	2.269	1.678-3.068	0.819	< 0.001
High AST	1.246	0.776-2.002	0.220	0.362

Age was excluded due to poor model fit as indicated by the Hosmer-Lemeshow test.

BMI: Body mass index; FAST: FibroScan-AST; LDL: Low-density cholesterol; HDL: High-density cholesterol; TGC: Triglycerides; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; GGT: Gamma-glutamyl transferase; HbA1c: Glycated hemoglobin; OR: Odds ratio; CI: Confidence interval.

The multivariate analysis identified BMI class, triglycerides, HbA1c ≥ 6.5%, and elevated ALT (as defined by guideline cutoffs) as the main factors statistically associated with the presence of liver steatosis.

Independent risk factors for the presence of clinically significant liver fibrosis: To identify independent risk factors associated with the presence of clinically significant liver fibrosis (LSM ≥ 8kPa), a logistic regression analysis was conducted, as shown in Table 5. The final model identified BMI class, the presence of liver steatosis (as defined by CAP), GGT levels, and elevated ALT and AST (according to guideline cutoffs) as the main predictors of liver fibrosis in the cohort. Among these, elevated transaminases showed the highest odds ratios.

Table 5 Logistic regression to evaluate the risk factors associated with presence of liver fibrosis.

Univariate	OR	95%CI	β	P value
Age	1.006	0.988-1.024	0.006	0.508
BMI class	1.783	1.492-2.131	0.578	< 0.001
Sex	0.809	0.551-1.187	-0.212	0.278
HbA1c ≥ 6.5%	1.664	1.074-2.580	0.509	0.023
Steatosis (CAP)	2.995	2.060-4.355	1.097	< 0.001
Triglycerides mg/dL	1.000	0.998-1.002	0.000	0.816
LDL cholesterol	0.994	0.988-1.001	-0.006	0.073
High AST	12.925	8.121-20.572	2.559	< 0.001
High ALT	6.681	4.160-10.729	1.899	< 0.001
GGT	1.026	1.019-1.032	0.025	< 0.001
Multivariate
BMI Class	1.500	1.169-1.924	0.405	0.001
HbA1c ≥ 6.5%	1.309	0.765-2.242	0.269	0.326
Steatosis (CAP)	1.965	1.162-3.324	0.676	0.012
High AST	4.180	2.314-7.549	1.430	< 0.001
High ALT	2.193	1.184-4.061	0.785	0.013
GGT	1.021	1.013-1.028	0.020	< 0.001

BMI: Body mass index; FAST: FibroScan-AST; LDL: Low-density cholesterol; HDL: High-density cholesterol; TGC: Triglycerides; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; GGT: Gamma-glutamyl transferase; HbA1c: Glycated hemoglobin; CAP: Controlled attenuation parameter; OR: Odds ratio; CI: Confidence interval.

Supplementary Table 4 shows a sensitivity analysis of clinically significant fibrosis using alternative liver stiffness thresholds.

Prevalence of altered transaminases and GGT levels in the population

We evaluated the prevalence of altered transaminase levels based on guideline-recommended thresholds: 33 U/L for ALT in men and 25 U/L in women[17]. Elevated ALT levels were found in 32.6% of men, 41% of women, and 37.6%of the total population. AST levels exceeded the reference range (13-39 U/L) in 9.5% of the population, while GGT levels were above the local laboratory reference range (6-64 U/L) in 9.1%.

Overall, 40.6% of patients had abnormal ALT, AST, or GGT levels. Among them, 9.8% had abnormalities in two markers, while 3.0% had abnormalities in all three.

Risk of MASLD with significant activity and fibrosis (at-risk MASH)

Finally, 2.2% of the total population had a FAST score ≥ 0.67, indicating a high risk of MASLD with significant activity and fibrosis (MASH). After adjusting the FAST score according to HbA1c levels (≥ 0.67 for HbA1c < 6.5%and 0.5 for HbA1c ≥ 6.5%), the prevalence increased to 4.8%. Obesity, ALT, and GGT levels were associated with at-risk MASH according to HbA1c levels. Logistic regression analysis identified obesity, ALT, and GGT as independently linked to at-risk MASH (Supplementary Table 5).

DISCUSSION

In this study, we describe the risk factors associated with the presence of steatosis, liver fibrosis, and at-risk MASH in 1337 patients recently diagnosed with T2DM (less than 5 years) from the Mexican population, which is known for its high susceptibility to metabolic diseases, including MASLD. The main findings of the study are shown in Figure 1.

Figure 1 Main findings of the study. MASLD: Metabolic dysfunction-associated steatotic liver disease; T2DM: Type 2 diabetes mellitus.

Our study reported a prevalence of MASLD of 41.4%, similar to that reported in the literature[18,19], and a high prevalence of overweight/obesity, consistent with recent findings from the ENSANUT 2022 (National Survey of Health and Nutrition). One important consideration contrasting our results with previous ones is the reported prevalence. A recent meta-analysis in T2DM reported an overall NAFLD prevalence of 65.3%, using any imaging method, including magnetic resonance imaging, USG, VCTE/CAP, serum biomarkers, and computed tomography. When specifically addressing those studies with CAP, NAFLD prevalence was 45.8% (27.05%-65.85%) for CAP of 275-302 dB/m, and 80.64% (75.27%-85.08%) for CAP 222-249 dB/m (P < 0.001 for the difference between CAP groups)[20]. Using a higher CAP (≥ 288 dB/m) as suggested by the current guidelines, as well as the fact of having good metabolic control (HbA1c 6.8%), and an earlier onset of T2DM, explain the lower prevalence of MASLD. When we used a CAP cutoff of > 250 dB/m[21], MASLD prevalence increased to 66%.

As previously described, patients with liver steatosis had a higher prevalence of metabolic abnormalities, liver chemistry alterations, and elevated HbA1c levels. A key finding of our study was the high prevalence of clinically significant liver fibrosis (≥ F2), estimated at 10% in this population by VCTE. Although this percentage is lower than that reported in other studies[4,5], this could be attributed to the early stages of the disease in the patients, the absence of macro/microvascular complications, and the fact that most participants maintained good metabolic control (with an average HbA1c of 6.8% for the overall population, and more than half having values < 7%).

Regarding the duration of T2DM, although our population was well characterized to ensure early diagnosis, a potential diagnostic delay (a frequent and challenging problem in this disease) may be present. The onset of T2DM has been estimated to be 4-6 years before a clinical diagnosis[22], allowing untreated and persistent hyperglycemia to possibly exert a negative effect on the liver and, therefore, likely affecting the data on the prevalence of liver steatosis and fibrosis.

Remarkably, 56 patients had a high risk of advanced fibrosis, with 40 already at stage F4. Up to 4.8% of the population was categorized as at risk for MASH according to the FAST score. None of these patients had a previous liver diagnosis and were unaware of their chronic liver disease. The overall prevalence of liver fibrosis can be influenced by the different LSM cutoffs reported in the literature, with higher values inherently increasing specificity and decreasing sensitivity (a sensitivity analysis of the present study is shown in Supplementary Table 4). However, in the present work, a cutoff of 8 kPa was set as the initial value for risk of clinically significant fibrosis (F ≥ 2), as recommended by the AASLD[2], EASL-EASD-EASO[23] and ADA[24] to provide a more widely used and pragmatic value for clinical practice. Positive predictive value (PPV) for FAST ≥ 0.67 was 0.83 and 0.69 for the derivation and validation cohorts in the original manuscript developing the FAST score[14]. On the other hand, the use of the AASLD clinical pathway for fibrosis-risk stratification in T2DM using VCTE has shown a low false-negative rate (3.3%)[25]. Specifically in T2DM, VCTE values of < 8 kPa and > 12.0 kPa have shown a PPV of 67.8%, and 77.9%, respectively[26].

As with patients with steatosis, those with clinically significant fibrosis had higher BMI and altered blood chemistry, particularly elevated HbA1c, AST, and GGT concentrations.

When stratifying the population by HbA1c levels (≥ 7% vs < 7%), we observed higher CAP values in patients with HbA1c > 7% (P = 0.040). Similarly, patients with HbA1c > 7% had elevated GGT levels, suggesting hepatocyte alterations induced by metabolic stress. This trend was more pronounced across HbA1c groups, with a stepwise increase in GGT levels. Interestingly, logistic regression analysis showed that GGT levels were associated with at-risk MASH.

Liver fibrosis stands as the most crucial prognostic factor for both liver-and non-liver-related outcomes, including mortality in MASLD patients. It has been reported that, on average, fibrosis progresses by one stage every 7 years; however, in patients with T2DM, this progression is faster[19], likely due to increased metabolic and oxidative stress[27]. Our study identified BMI, ALT, AST, and GGT as independent factors associated with liver fibrosis, suggesting an important role of metabolic stress as a driver of fibrosis response in this population.

These findings highlight the importance of detecting both liver steatosis and, more critically, liver fibrosis in patients with T2DM. Early identification of MASLD in these patients is crucial for the timely intensification of metabolic control. Additionally, patients with fibrosis should be considered for referral to specialized centers evaluating new targeted treatments for both steatosis and fibrosis and for liver disease staging and follow-up.

Our study highlights the importance of basic liver chemistry tests, such as ALT, AST, and GGT levels. This is crucial given the high prevalence of abnormalities, even among recently diagnosed T2DM patients (40.6% of the population), and the potential to identify patients at risk of MASH when combined with transient elastography.

This study has several strengths, including a robust sample size, a well-defined population of T2DM patients, appropriate methodology for VCTE examinations, and the use of established cut-off points for diagnosing liver fibrosis, steatosis, and at-risk MASH. However, it also has limitations, including the lack of recent HbA1c values for a subset of patients, the absence of liver biopsy data, and the potential for FAST score and VCTE measurements to be influenced by T2DM and obesity. Additionally, platelet data were not available as they are not part of the regular assessment at the diabetes clinic. As a result, some non-invasive risk scores could not be calculated. Although combining non-invasive risk scores with TE results or a stepwise evaluation could have been beneficial, we believe the direct transient elastography measurement provides a robust approach in this high-risk population.

Although the present study was conducted in a highly structured and specialized diabetes care center (a single tertiary referral center), the results clearly underscore the importance of screening patients with relatively good metabolic control and in the early stages of the disease. Although the generalizability of the findings in this study may be limited in primary care settings without access to VCTE, the high prevalence of MASLD with fibrosis makes liver evaluation mandatory (even with blood tests such as ALT, AST, and GGT) in patients with T2DM, even in early stages of the disease. Therefore, we recommend routinely evaluating patients with early-diagnosis T2DM using a stepwise approach with blood-based liver fibrosis scores and VCTE[28].

Beyond highlighting the high prevalence of steatosis, fibrosis, and at-risk MASH, this study underscores the importance of screening in high-risk populations, such as the Mexican population, where 18.3% of adults have T2DM. Given that 66.2% of the country's population (126 million people) are adults (INEGI), the estimated number of individuals with both MASLD and T2DM is striking, exceeding 15 million. One important consideration is that these numbers represent only adults with T2DM. Therefore, within the total population of the country, including individuals with other risk factors such as overweight, obesity, and dyslipidemia, the prevalence is likely even higher.

In addition to diagnosing and stratifying MASLD in patients, the high prevalence of MASLD in diabetes calls for including liver health evaluations in all T2DM patients. It also highlights the need to explore new interventions, both pharmacological and non-pharmacological, to address metabolic derangements, liver steatosis, and fibrosis.

CONCLUSION

The prevalence of steatosis and clinically significant liver fibrosis among patients recently diagnosed with T2DM (less than 5 years) was 41.4% and 10%, respectively. Additionally, 3% of the population had cirrhosis. Risk factors independently associated with these conditions include higher BMI, transaminase levels, GGT, triglycerides, and HbA1c. Therefore, routine assessment of these factors and liver imaging is necessary in the T2DM population to detect MASLD.

ACKNOWLEDGEMENTS

We thank our interns and technicians for their technical support. JAGR was supported by Fundación Mexicana para la Salud Hepática (FUNDHEPA) through the Estímulo Ángeles Espinoza Yglesias award 2022.