Changes in lipid biomarkers, liver stiffness and steatosis following sodium-glucose cotransporter two inhibitor therapy in type two diabetes

Abstract

BACKGROUND

Metabolic dysfunction-associated steatotic liver disease is a rapidly growing and often underrecognized public health challenge affecting most patients with type 2 diabetes mellitus (T2DM). Despite its high prevalence and substantial hepatic and cardiovascular burden, effective pharmacological treatment remains limited. Sodium-glucose cotransporter 2 inhibitors (SGLT2i) are promising therapeutic candidates because they lower glucose independently of insulin, improve insulin sensitivity, and provide renoprotective and cardioprotective benefits, with emerging evidence of potential hepatoprotective effects.

AIM

To assess changes in circulating lipogenesis biomarkers and liver imaging parameters following SGLT2i therapy.

METHODS

In this single-arm prospective observational study, 79 patients with T2DM initiating SGLT2i therapy were followed for six months. Peroxisome proliferator-activated receptor alpha (PPAR-α) and peroxisome proliferator-activated receptor gamma (PPAR-γ), sterol regulatory element-binding protein-1c (SREBP-1c), and microsomal triglyceride transfer protein (MTTP) were assessed by enzyme-linked immunosorbent assay. Liver stiffness and steatosis were assessed using two-dimensional shear wave elastography (2D-SWE) and ultrasound-guided attenuation parameter (UGAP), with additional biochemical, anthropometric, and noninvasive fibrosis assessments.

RESULTS

Sixty-seven patients with T2DM completed six months of SGLT2i therapy. Treatment was associated with significant reductions in body mass index, waist circumference, fasting plasma glucose, and glycated hemoglobin (all P ≤ 0.04). Circulating PPAR-γ levels decreased from 895 ng/L to 565.4 ng/L, whereas MTTP concentrations increased from 198.4 ng/L to 428.9 ng/L (P < 0.001 for both). 2D-SWE-measured liver stiffness decreased from 6.2 kPa to 6.1 kPa (P < 0.001) and correlated positively with UGAP-assessed liver steatosis (ρ = 0.422, P < 0.001) and inversely with MTTP levels (ρ = -0.268, P = 0.03). Erythrocyte count, hemoglobin, hematocrit, and red blood cell indices changed significantly, while MTTP, SREBP-1c, and PPAR-α showed distinct associations with lipid and liver function markers (all P ≤ 0.04). Fibrosis and steatosis stages, along with non-alcoholic fatty liver disease fibrosis score and Fibrosis-4 index risk categories, remained largely stable, with a numerical increase in F0/F1 fibrosis, a decrease in F4, and balanced steatosis stages.

CONCLUSION

SGLT2i therapy was associated with modulation of circulating PPAR-γ and MTTP levels and a statistically significant but modest reduction in liver stiffness whose clinical relevance remains uncertain, while fibrosis and steatosis categories remained largely stable over 6 months. These findings support the metabolic effects of SGLT2i and warrant confirmation in larger controlled studies with longer follow-up and histological endpoints to determine whether these effects are accompanied by improvements in liver fibrosis and steatosis.

Key Words: Metabolic dysfunction-associated steatotic liver disease; Type 2 diabetes mellitus; Sodium-glucose cotransporter 2 inhibitors; Two-dimensional shear wave elastography; Ultrasound-guided attenuation parameter; Sterol regulatory element-binding protein 1c; Peroxisome proliferator-activated receptor gamma; Peroxisome proliferator-activated receptor alpha; Microsomal triglyceride transfer protein

Core Tip: Metabolic dysfunction-associated steatotic liver disease lacks effective pharmacologic therapies and robust biomarkers for monitoring treatment response. In this prospective observational study, sodium-glucose cotransporter 2 inhibitors therapy was associated with modulation of selected circulating lipogenesis biomarkers and a modest reduction in liver stiffness of uncertain clinical relevance, whereas steatosis measures remained largely unchanged over 6 months.

INTRODUCTION

Metabolic dysfunction-associated steatotic liver disease (MASLD), previously termed non-alcoholic fatty liver disease (NAFLD), affects 30%-40% of adults, 60%-70% of patients with type 2 diabetes mellitus (T2DM), and up to 80% of individuals with obesity, and is a leading cause of liver transplantation in Europe and the United States[1,2]. Diagnosis of MASLD requires evidence of hepatic steatosis on imaging or histology exceeding 5% of liver weight, the presence of at least one cardiometabolic risk factor such as obesity, T2DM, hypertension, or dyslipidemia, and exclusion of other causes of hepatic steatosis, including clinically significant alcohol consumption or other liver diseases[3,4]. MASLD ranges from metabolic dysfunction-associated steatotic liver to metabolic dysfunction-associated steatohepatitis (MASH), which can progress to severe fibrosis, cirrhosis, and hepatocellular carcinoma[3]. It is closely linked to extrahepatic complications, particularly cardiovascular disease, which constitutes the primary cause of mortality and escalates with advancing fibrosis severity[2].

Because MASLD is often asymptomatic in its early stages, systematic screening of individuals at risk is critical[3]. Liver biopsy remains the reference standard for assessing liver histology, but it is ethically impractical for routine screening due to procedural risks, sampling variability and cost[5]. Conventional noninvasive imaging modalities for assessing liver fibrosis and steatosis include ultrasound, computed tomography, and magnetic resonance imaging. Among these, ultrasound is widely available, safe, and cost-effective, and provides immediate real-time results, but is operator dependent and its sensitivity for detecting mild steatosis is limited, particularly in advanced fibrosis[6,7]. Novel ultrasound techniques, two-dimensional shear wave elastography (2D-SWE) and the ultrasound-guided attenuation parameter (UGAP), improve diagnostic accuracy and enable comprehensive quantitative assessment of hepatic fibrosis and steatosis, supporting routine screening in MASLD[8]. Specifically, 2D-SWE evaluates tissue stiffness by measuring shear wave velocity, with higher velocities reflecting fibrosis, whereas UGAP quantifies hepatic steatosis by assessing ultrasound attenuation, which increases with liver fat content[7-9]. In addition, noninvasive fibrosis assessments using the Fibrosis-4 index (FIB-4) and NAFLD Fibrosis Score (NFS) algorithm, a validated tools based on routine clinical and laboratory parameters, enable risk stratification for advanced fibrosis and complement imaging-based evaluation[10,11].

Recently FDA-approved agents, the thyromimetic resmetirom, and the glucagon-like peptide-1 (GLP-1) receptor agonist semaglutide, improve steatosis, inflammation, and fibrosis, however, their clinical applicability remains constrained by restricted global accessibility, incompletely characterized metabolic effects on hepatic pathways, and uncertain long-term outcomes. Consequently, pharmacologic options remain insufficient to fully address the metabolic and hepatic complexity of MASLD[12,13]. Sodium-glucose cotransporter 2 inhibitors (SGLT2i), or “flozins”, are oral antidiabetic agents initially approved for T2DM and later for chronic kidney disease and heart failure. SGLT2i are promising candidates for MASLD therapy because they lower blood glucose independently of insulin by increasing urinary glucose excretion, reduce hyperinsulinemia, improve insulin sensitivity, and confer renoprotective and cardioprotective effects, including improved cardiac energy metabolism, reduced oxidative stress, and modulation of blood pressure and glomerular function[14,15]. Preclinical studies also suggest direct hepatoprotective effects, including reduced hepatic lipid accumulation, attenuation of inflammation, and modulation of hepatocyte proliferation and apoptosis[14,16].

In MASLD, dysregulation of key lipogenesis biomarkers contributes to hepatic fat accumulation and liver injury, making them important therapeutic targets and potential indicators of treatment response[17]. Sterol regulatory element-binding protein-1c (SREBP-1c) and peroxisome proliferator-activated receptor gamma (PPAR-γ) are typically overexpressed in hepatic tissue, promoting triglyceride synthesis and lipid storage, whereas peroxisome proliferator-activated receptor alpha (PPAR-α) and microsomal triglyceride transfer protein (MTTP) are downregulated, impairing fatty acid oxidation and lipoprotein export. SGLT2i positively modulate these pathways in animal models, enhancing β-oxidation, reducing hepatic lipid accumulation, and improving insulin sensitivity, although clinical evidence in patients remains limited[14,17-21]. Because liver tissue is not routinely available in clinical practice, circulating (serum) lipogenesis-related biomarkers offer a practical and minimally invasive approach for investigating these metabolic pathways[22]. However, whether their circulating concentrations accurately reflect intrahepatic gene or protein expression remains uncertain. Therefore, they should be interpreted as indirect indicators of metabolic processes rather than as validated markers for MASLD assessment.

Given the high prevalence of MASLD among patients with T2DM and the emerging hepatometabolic effects of SGLT2i, this study evaluated changes in circulating lipogenesis biomarkers and ultrasound-derived liver parameters following SGLT2i therapy in patients with T2DM to provide insights into hepatometabolic mechanisms potentially relevant to MASLD. We hypothesized that SGLT2i therapy modulates lipogenesis-related biomarkers and liver stiffness. The primary objectives were to measure serum SREBP-1c, PPAR-α, PPAR-γ, and MTTP, and to assess liver stiffness using 2D-SWE after six months of therapy. Secondary objectives included assessment of hepatic steatosis by UGAP, standard biochemical markers of liver injury and lipid metabolism and correlations among molecular biomarkers, imaging findings, biochemical parameters and the FIB-4 and NFS. Lifestyle and dietary data were collected for participant monitoring and verification of continued eligibility.

MATERIALS AND METHODS

Study design and ethical considerations

This prospective observational cohort study was conducted from March 2024, when the first participant was enrolled, to October 2025, when the last participant completed the six-month follow-up. Assessments were performed at two time points: Before initiation of SGLT2i therapy and six months after treatment. The primary endpoint was the change in liver stiffness measured by 2D-SWE and in selected lipogenesis biomarkers after 6 months of therapy; secondary endpoints included changes in UGAP, biochemical markers, anthropometric measures, and noninvasive fibrosis scores. Ethical approval was obtained from the Ethical Committee of the Faculty of Medicine, Josip Juraj Strossmayer University of Osijek (Approval No. 2158-61-46-23-122), the Health Center of Osijek-Baranja County (Approval No. 03-1311-1/23), the General County Hospital Našice (Approval No. 01-348/2-2024), and the Faculty of Dental Medicine and Health Osijek, Josip Juraj Strossmayer University of Osijek (Approval No. 2158/97-97-10-24-55). The study was conducted in accordance with the Declaration of Helsinki, and all participants provided written informed consent. The trial was registered at ClinicalTrials.gov (ID: NCT07269197).

Participants

This study included a total of 67 adults with T2DM who were initiating SGLT2i therapy for the first time at the Department of Internal Medicine, General County Hospital Našice, and at general medical practices in Osijek and surrounding areas (Table 1). The minimum required sample size was 57 participants, as determined by an a priori power analysis performed using G*Power software (version 3.1.2) to detect a medium effect size (Cohen’s d = 0.5) in paired comparisons of continuous variables between baseline and 6-month follow-up, with a significance level of 0.05 and statistical power of 95%. To account for an anticipated 10% attrition rate during follow-up, the number of participants planned for initial enrollment was increased to 63. Eligible participants were 18 years or older, had a diagnosis of T2DM, provided written informed consent, were receiving an SGLT2i for the first time, and had been on stable antihyperglycemic therapy for at least three months prior to enrollment. Patients were excluded if they were on insulin therapy, taking highly hepatotoxic drugs or medications associated with drug-induced fatty liver disease, including amiodarone, tamoxifen, methotrexate, 5-fluorouracil, irinotecan, l-asparaginase, valproate, tetracycline, or nucleoside reverse transcriptase inhibitors, or if they had liver cancer, cirrhosis, hemochromatosis, autoimmune or cholestatic liver disease, or hepatitis B or C infection. Additional exclusion criteria were alcohol dependence or excessive alcohol consumption, pregnancy or breastfeeding, or mental incapacity requiring a legal guardian.

Table 1 Basic sociodemographic characteristics and comorbidities of study participants (n = 67).

Characteristic		n (%)
	Median age and range in years	66 (37-89)
Gender	Male	40 (59.7)
	Female	27 (40.3)
Duration of diabetes mellitus type 2 (T2DM)	< 5 years	40 (59.7)
	5-10 years	16 (23.9)
	11-15 years	5 (7.5)
	> 15 years	6 (9.0)
Comorbidities (other than T2DM)	Hypertension	53 (79)
	Kidney disease	9 (13.4)
	Hyperlipidemia	42 (62.7)
	Hypothyroidism	3 (4.5)
	Hyperthyroidism	1 (1.5)
	Depression	4 (6.0)
	Anxiety	2 (3.0)
	Liver disease	6 (9.0)
	Gallbladder and bile duct disease	5 (7.5)
Other comorbidities		14 (20.9)

Imaging assessment

Liver stiffness measurement was performed using the 2D-SWE and liver steatosis was assessed using the UGAP at the Health Center Osijek on a LOGIQ P9 ultrasound system (GE HealthCare, Chicago, Illinois, United States)[23]. Patients fasted for at least four hours and were positioned supine with the right arm abducted. The right liver lobe was accessed through the intercostal spaces in B-mode. For 2D-SWE, once a high-quality image of the liver parenchyma was obtained, patients were instructed to inhale lightly and hold their breath in a neutral position for 3-5 seconds. A stabilized elastogram was acquired and the region of interest (Q-box) was positioned within a homogeneous area. Up to 10 measurements were recorded for each patient, and the median was reported as the final liver stiffness in kPa[24,25]. Subsequently, the system was switched to UGAP mode, and a homogeneous region of the liver parenchyma, free of large vessels, was identified. Once located, the image was frozen and multiple measurements could be performed within the area. Up to ten readings were obtained per patient and the median was reported as the final steatosis value in decibels per centimeter per megahertz (dB/cm/MHz)[26]. All 2D-SWE and UGAP measurements were performed by a single experienced operator who was blinded to the laboratory results. Measurement reliability was assessed using the interquartile range-to-median ratio, with values below 30% considered acceptable for both 2D-SWE and UGAP measurements[27,28]. Patients were categorized by fibrosis stage using the following 2D-SWE cutoff values: F0-F1 (≤ 8.26 kPa), F2 (8.27-9.39 kPa), F3 (9.40-11.87 kPa), and F4 (≥ 11.88 kPa)[29]. Steatosis was graded using the following UGAP cutoff values: S0 (≤ 0.64 dB/cm/MHz), S1 (0.65-0.70 dB/cm/MHz), S2 (0.71-0.76 dB/cm/MHz), and S3 (≥ 0.77 dB/cm/MHz)[26].

Laboratory analysis

Blood samples were collected from all patients at the Department of Internal Medicine, General County Hospital Našice, or general medical practices in Osijek-Baranja County. Standard laboratory analyses included fasting glucose, glycated hemoglobin (HbA1c), complete blood count, alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase, gamma-glutamyl transferase (GGT), albumin, total protein, coagulation parameters, total and direct bilirubin, and lipid profile. Serum lipogenesis-related biomarkers, SREBP-1c, PPAR-α, PPAR-γ, and MTTP, were assessed by enzyme-linked immunosorbent assay (ELISA) at the Laboratory for Translational Medicine of the Faculty of Dental Medicine and Health Osijek, Josip Juraj Strossmayer University of Osijek. ELISA for lipogenesis biomarkers was performed using commercial sandwich kits (BT LAB, Korain Biotech, China) according to the manufacturer’s instructions. The following kits were used: PPAR-α (Cat. No. E1450Hu; detection range 30-9000 ng/L; intra-assay coefficient of variation (CV) 3.4%-4.8%; inter-assay CV < 10%, PPAR-γ (Cat. No. E1511Hu; detection range 30-9000 ng/L; intra-assay CV 4.0%-7.6%; inter-assay CV < 10%), SREBP-1c (Cat. No. E6666 Hu; detection range 200-60000 ng/L; intra-assay CV 3.4%-5.6%; inter-assay CV < 10%), and MTTP (Cat. No. E5673 Hu; detection range 15-3000 ng/L; intra-assay CV 3.34%-5.79%; inter-assay CV < 10%). Serum samples were incubated on antibody-coated plates with a biotinylated detection antibody and Streptavidin-horseradish peroxidase, followed by substrate addition and reaction termination. Absorbance was read at 450 nm, and biomarker concentrations were calculated using standard curves.

Lifestyle and dietary assessment

Lifestyle and dietary data were collected using a study-specific questionnaire at baseline and at the 6-month follow-up visit. The questionnaire was used to assess selected lifestyle-related characteristics and monitor continued compliance with the study inclusion and exclusion criteria throughout follow-up. Participants were instructed to accurately report any changes in lifestyle-related characteristics, including alcohol consumption, smoking status, physical activity, meal frequency, and dietary supplement use, to allow comparison between baseline and follow-up. Furthermore, participants who no longer fulfilled the predefined eligibility criteria were excluded from the final analysis. Investigator-measured variables included weight, height for body mass index (BMI), blood pressure, and waist circumference. Participants also self-reported education, medical history, recent antibiotic use, pregnancy or breastfeeding, treatment adherence, and concomitant medication use.

Noninvasive fibrosis assessment

To enable noninvasive stratification of advanced fibrosis risk alongside imaging-based evaluation, we calculated FIB-4 and NFS. FIB-4 incorporates age, aminotransferase levels (AST, ALT), and platelet count, whereas NFS applies a multivariable algorithm including age, BMI, diabetes status, AST/ALT ratio, platelet count, and serum albumin.

The FIB-4 was calculated as: FIB-4 = age × AST/[platelet count × √(ALT)] (age in years, ALT and AST in U/L, and platelet count in 10⁹/L).

The NFS was calculated as: NFS = -1.675 + 0.037 × age + 0.094 × BMI + 1.13 × impaired fasting glucose (yes = 1, no = 0) + 0.99 × AST/ALT - 0.013 × platelet count - 0.66 × albumin (age in years; BMI in kg/m²; AST, ALT in U/L; platelet count in 10⁹/L and albumin in g/dL).

Based on the calculated FIB-4 and NFS values, patients were stratified into three categories reflecting low, intermediate, and high risk for advanced fibrosis. The NFS cutoff values were defined as < -1.5 (low risk), -1.5 to < 0.67 (intermediate risk), and ≥ 0.67 (high risk). For FIB-4, values < 1.3, 1.3-2.67, and > 2.67 indicated low, intermediate, and high risk, respectively, in patients aged ≤ 65 years. In patients aged > 65 years, the lower cutoff was increased to 2.0 to reduce false-positive classifications; values < 2.0, 2.0-2.67, and > 2.67 indicated low, intermediate, and high risk, respectively[10,11].

Statistical analysis

Categorical data were presented as absolute and relative frequencies. Differences in dependent categorical variables were tested using the McNemar-Bowker test or the test of marginal homogeneity. The normality of numerical variables was assessed using the Shapiro-Wilk test. Given the distribution of the data, non-parametric methods were applied. Continuous data were described using the median and interquartile range. Differences between measurements were tested using the Wilcoxon signed-rank test, with the reported difference and the Hodges-Lehmann estimate of the median difference. Differences between three or more independent groups were assessed using the Kruskal-Wallis test (with post hoc Conover test). Associations between variables were evaluated using Spearman’s rank correlation coefficient[30,31]. All tests were two-tailed with a significance level of α = 0.05. Statistical analyses were performed using MedCalc^® Statistical Software version 23.4.8 (MedCalc Software Ltd, Ostend, Belgium; 2026) and SPSS Statistics for Windows, version 25.0 (IBM Corp., Armonk, NY, United States).

RESULTS

Participant flow and characteristics

A total of 84 patients with T2DM were assessed for eligibility, and 17 were excluded as shown in the flow diagram (Figure 1). Reasons for study exclusion varied, including one patient who discontinued SGLT2i due to genitourinary adverse effects. At the 6-month follow-up visit, five participants were excluded from the final analysis after questionnaire assessments identified changes affecting study eligibility: Three because of poor adherence to SGLT2i therapy, one because of excessive alcohol consumption, and one because insulin therapy was initiated. The final cohort therefore included 67 patients (40 men and 27 women), aged 37 years to 89 years, with a median age of 66 years and a median BMI of 31.46 kg/m².

Figure 1 Flow diagram showing patient screening, enrollment, exclusions, and study completion. A total of 84 patients were assessed for eligibility, 79 were enrolled, and 67 completed the study. SGLT2i: Sodium-glucose cotransporter 2 inhibitor.

Most participants had been diagnosed with T2DM within the past five years. Empagliflozin was prescribed to 29 patients as either empagliflozin monotherapy or the fixed-dose combination empagliflozin/metformin, resulting in daily empagliflozin doses of 5 mg (n = 1), 10 mg (n = 7), 12.5 mg (n = 16), and 25 mg (n = 5). Dapagliflozin was prescribed to 38 patients as either dapagliflozin monotherapy or the fixed-dose combination dapagliflozin/metformin, resulting in daily dapagliflozin doses of 5 mg (n = 1) and 10 mg (n = 37). Fixed-dose combinations were prescribed only to patients who had already been receiving an established metformin regimen before enrollment, which remained unchanged throughout the study. Hyperglycemia was present in all participants. Hypertension was observed in 53 participants, hyperlipidemia in 42, and chronic kidney disease in 9 (Table 1). Fourteen participants reported a history of liver-related conditions, most commonly fatty liver or hepatic steatosis. Most participants were on antidiabetic medication (n = 66, 98.5%), followed by antihypertensive (n = 48, 71.6%) and lipid-lowering therapy (n = 32, 47.8%). Cardiac medications (n = 16, 23.9%), anticoagulant or antiplatelet therapy (n = 11, 16.4%), and diuretics (n = 9, 13.4%) were less common. Fewer participants used medications for thyroid disorders, depression, or anxiety, and 9 participants reported other concomitant medications. At baseline, approximately half of the participants reported occasional alcohol consumption, 22.7% (n = 15) were current smokers, and 15.0% (n = 10) reported dietary supplement use. Paired analyses of alcohol consumption, smoking status, physical activity, meal frequency, and dietary supplement use showed no statistically significant changes over the 6-month follow-up (all P > 0.05; Table 2).

Table 2 Changes in alcohol consumption, smoking status, and lifestyle characteristics of the study participants between baseline and 6 months of sodium-glucose cotransporter 2 inhibitor therapy.

Parameter		n (%) or n/N1		P value2
		Baseline	6 months
Frequency of alcohol consumption	No alcohol consumption	31 (46.3)	29 (43.9)	0.36
	Occasional alcohol consumption	32 (47.8)	31 (47.0)
	Regular alcohol consumption	4 (6.0)	6 (9.1)
1Regular drinkers-frequency (N = 4) (e.g., 250 mL of beer, 125 mL of wine or 30 mL of spirits)	Up to 2 drinks per week	1/4	1/4	0.323
	Up to 2 drinks per day	2/4	3/4
	Up to 3 drinks per day	1/4	0
Cigarette smoking status	Never smoker	28 (42.4)	23 (36.5)	0.25
	Former smoker	23 (34.8)	24 (38.1)
	Current smoker	15 (22.7)	16 (25.4)
1Daily cigarette consumption among smokers (N = 15)	≤ 20 cigarettes/day	5/15	7/15	0.183
	21-40 cigarettes/day	9/15	8/15
	> 40 cigarettes/day	1/15	0
1Duration of smoking (N = 15)	1–10 years	1/15	1/15	> 0.99
	More than 20 years	14/15	14/15
Main daily activity lasting more than 6 hours (N = 65)	Sitting	30 (46.2)	30 (40.2)	0.90
	Standing	10 (15.4)	8 (12.3)
	Walking	26 (38.5)	27 (41.5)
Number of main meals consumed per day (N = 66)	One	4 (6.1)	3 (4.5)	0.51
	Two	25 (37.9)	21 (31.8)
	Three	30 (45.5)	38 (57.6)
	Four or more	7 (10.6)	4 (6.1)
Dietary supplement use		10 (14.9)	16 (26.7)	0.06

¹n denotes the number of participants, and N denotes the total number of participants in the relevant subgroup (current smokers or regular drinkers).

²McNemar-Bowker’s test.

³Marginal Homogeneity test.

Effect of 6-month SGLT2i therapy on biochemical parameters, lipogenesis biomarkers and liver histology

After six months of SGLT2i therapy, median BMI decreased from 31.46 kg/m² to 30.12 kg/m² and waist circumference from 107 cm to 104 cm, both P < 0.001. Glycemic parameters improved, with fasting plasma glucose declining to 6.4 mmol/L (P = 0.006) and HbA1c to 6.7 percent (P = 0.008). Serum total protein and total bilirubin increased significantly (P ≤ 0.02). Modest changes were observed in coagulation parameters, including prothrombin time ratio, and international normalized ratio (P ≤ 0.002), with fibrinogen activity showing a statistical trend (P = 0.05). Erythrocyte count, hemoglobin, and hematocrit increased significantly compared with baseline (all P < 0.001), whereas liver enzymes and lipid parameters remained stable (Table 3).

Table 3 Changes in biochemical and anthropometric parameters from baseline to 6 months after initiation of sodium-glucose cotransporter 2 inhibitor therapy.

Parameter	Baseline, median (IQR)	6 months, median (IQR)	Difference1	95%CI	P value2
BMI (kg/m2)	31.46 (28.35-34.67)	30.12 (28.09-34.34)	-0.79	-1.1 to -0.5	< 0.001a
WC (cm)	107 (102-115)	104 (97.75-107)	-4	-5 to -2.5	< 0.001a
PG (mmol/L)	6.9 (6.3-8.6)	6.4 (5.5-7.6)	-0.6	-1 to -0.2	0.006a
HbA1c (%)	7 (6.3-8)	6.7 (6.3-7.3)	-0.25	-0.5 to -0.05	0.008a
HbA1c (mmol/mol)	52.5 (44.3-63.8)	50 (45-56)	-2	-4.5 to 0	0.04a
ALT (U/L)	28 (19-36)	26 (19-35)	-0.5	-3.5 to 1	0.41
AST (U/L)	25 (18-29)	24 (20-30)	0.5	-1.5 to 2	0.63
ALP (U/L)	67.5 (55-83.3)	67 (56.8-84.3)	0.5	-2 to 3.5	0.61
GGT (U/L)	26 (20-43)	25 (19-40)	-0.5	-2 to 1.5	0.68
Erythrocytes (1012/L)	4.7 (4.5-5.1)	5 (4.6-5.4)	0.24	0.15 to 0.32	< 0.001a
Hemoglobin (g/L)	141 (133-149)	146 (134-158)	4.5	2.5 to 6.5	< 0.001a
Hematocrit (L/L)	0.4 (0.4-0.4)	0.4 (0.4-0.5)	0.02	0.009 to 0.02	< 0.001a
MCV (fL)	88.3 (85.6-91.8)	88 (84.8-91.1)	-0.8	-1.5 to -0.2	0.02a
MCH (pg)	29.6 (28.7-30.5)	29.5 (27.8-30.8)	-0.4	-0.7 to -0.2	0.001a
MCHC (g/L)	335 (330-341)	334 (327-342)	-2	-4 to 0	0.06
RDW (%)	13.3 (12.7-14.1)	13.5 (13-14.5)	0.3	0.15 to 0.6	< 0.001a
Platelet count (109/L)	237 (202-270)	232 (193-264)	-3.5	-11.5 to 5	0.42
Leukocytes (109/L)	8 (6.3-9.3)	7.9 (6.9-9.6)	0.3	-0.05 to 0.55	0.09
Albumin (g/L)	44.7 (43-46.7)	45.3 (43.2-47.2)	0.3	-0.35 to 0.9	0.32
Total protein (g/L)	72 (69-75)	74 (70-77)	1.9	0.95 to 2.65	< 0.001a
Prothrombin time (ratio)	1.1 (1-1.2)	1.1 (1.1-1.2)	0.05	0.02 to 0.08	0.002a
Prothrombin time (INR)	1 (0.9-1.1)	1 (0.9-1)	-0.02	-0.05 to 0	0.001a
Fibrinogen-activity (g/L)	3.5 (2.9-4)	3.7 (3.1-4.3)	0.2	0 to 0.4	0.05
Thrombin time (second)	19.4 (18.4-21.5)	19.8 (19.2-20.8)	0.45	-0.4 to 1.3	0.33
Total bilirubin (µmol/L)	12 (9-14)	12 (9.8-16.3)	1.0	0 to 2	0.02a
Conjugated bilirubin (µmol/L)	2 (1-3)	2 (2-3)	0	0 to 0.5	0.27
Cholesterol (mmol/L)	4.4 (3.7-5.2)	4.2 (3.5-5.3)	0	-0.2 to 0.2	0.86
HDL (mmol/L)	1.2 (1-1.5)	1.3 (1.1-1.5)	0.05	-0.001 to 0.1	0.15
LDL (mmol/L)	2.4 (2-3.2)	2.3 (1.8-3.3)	-0.05	-0.25 to 0.1	0.42
Triglycerides (mmol/L)	1.9 (1.3-2.5)	1.8 (1.3-2.6)	0.03	-0.2 to 0.2	0.79

¹Hodges-Lehmann difference of median.

²Wilcoxon test.

^aP < 0.05.

ALT: Alanine aminotransferase; ALP: Alkaline phosphatase; AST: Aspartate aminotransferase; BMI: Body mass index; CI: Confidence interval; GGT: Gamma-glutamyl transferase; HbA1c: Glycated hemoglobin; HDL: High-density lipoprotein; INR: International normalized ratio; IQR: Interquartile range; LDL: Low-density lipoprotein; MCV: Mean corpuscular volume; MCH: Mean corpuscular hemoglobin; MCHC: Mean corpuscular hemoglobin concentration; RDW: Red cell distribution width; PG: Plasma glucose; WC: Waist circumference.

Significant alterations were observed in circulating lipogenesis-related biomarkers following treatment. PPAR-γ levels declined from a median of 895 ng/L to 565.4 ng/L (P < 0.001), whereas MTTP levels increased from a median of 198.4 ng/L to 428.9 ng/L (P < 0.001). In contrast, circulating PPAR-α and SREBP-1c levels did not change significantly from baseline to 6 months (PPAR-α, P = 0.38; SREBP-1c, P = 0.72; Table 4). Ultrasound imaging findings at the six-month evaluation showed that liver stiffness assessed by 2D-SWE decreased from a median of 6.2 kPa at baseline to 6.1 kPa at follow up (P < 0.001). Liver steatosis measured with UGAP showed a slight numerical decrease but did not reach statistical significance. Fibrosis and steatosis stages, as well as FIB-4 and NFS risk categories, remained largely stable after therapy, with a numerical increase in F0/F1 fibrosis and a decrease in F4 fibrosis compared with baseline (Table 5). Individual fibrosis stage transitions showed that 10 of 67 participants (14.9%) were reclassified to a lower fibrosis stage, 55 (82.1%) remained in the same stage, and 2 (3.0%) were reclassified to a higher fibrosis stage during follow-up. A representative elastogram from a single patient demonstrating a reduction in liver stiffness from stage F2 (8.31 kPa) to stage F0/F1 (6.94 kPa) after 6 months of SGLT2i therapy is shown in Figure 2.

Figure 2 Change in liver stiffness in a single patient following 6 months of sodium-glucose cotransporter 2 inhibitor therapy. A: Baseline liver stiffness corresponding to fibrosis stage F2 (8.31 kPa); B: Liver stiffness after 6 months of sodium-glucose cotransporter 2 inhibitor therapy corresponding to fibrosis stage F0/F1 (6.94 kPa).

Table 4 Changes in liver stiffness assessed by two-dimensional shear wave elastography, steatosis assessed by ultrasound-guided attenuation parameter, and lipogenesis biomarkers measured by enzyme-linked immunosorbent assay from baseline to 6 months after initiation of sodium-glucose cotransporter 2 inhibitor therapy.

Parameter	Baseline, median (IQR)	6 months, median (IQR)	Difference1	95%CI	P value2
Ultrasound imaging
E Median 2D-SWE (kPa)	6.2 (5.7-7.5)	6.1 (5.4-6.8)	-0.51	-0.93 to -0.21	< 0.001a
Med UGAP (dB/cm/MHz)	0.7 (0.7-0.8)	0.7 (0.6-0.8)	-0.02	-0.04 to 0.01	0.14
E IQR 2D-SWE (kPa)	1.3 (0.8-2.1)	0.1 (0-0.1)	-0.37	-0.69 to -0.09	0.005a
IQR UGAP (dB/cm/MHz)	0.05 (0.04-0.07)	0.06 (0.04-0.07)	0.01	-0.005 to 0.02	0.24
ELISA
PPAR-γ (ng/L)	895 (727.8-1105.6)	565.4 (336.7-839.8)	-297.03	-399.7 to -196.7	< 0.001a
PPAR-α (ng/L)	1667 (1133.7-3223.7)	1294 (779-2514)	-274.8	-818.5 to 308.7	0.38
SREBP-1c (ng/L)	5151.5 (3037.4-6969.7)	4900 (4100-5700)	174.3	-729.2 to 1062.0	0.72
MTTP (ng/L)	198.4 (103.8-301.3)	428.9 (351.6-532.1)	233.9	183.1 to 277.4	< 0.001a

¹Hodges-Lehmann difference of median.

²Wilcoxon test.

^aP < 0.001.

UGAP: Ultrasound-guided attenuation parameter; ELISA: Enzyme-linked immunosorbent assay; PPAR-γ: Peroxisome proliferator-activated receptor gamma; PPAR-α: Peroxisome proliferator-activated receptor alpha; SREBP-1c: Sterol regulatory element-binding protein-1c; MTTP: Microsomal triglyceride transfer protein; IQR: Interquartile range; CI: Confidence interval; 2D-SWE: Two-dimensional shear wave elastography.

Table 5 Distribution of participants by fibrosis stage, steatosis stage, nonalcoholic fatty liver disease fibrosis score and Fibrosis-4 index risk categories at baseline and after 6 months of sodium-glucose cotransporter 2 inhibitor therapy, n (%).

Parameter		Baseline	6 months	P value
Fibrosis stage	F0/F1	54 (80.6)	60 (89.6)	0.161
	F2	2 (3)	3 (4.5)
	F3	3 (4.5)	1 (1.5)
	F4	8 (11.9)	3 (4.5)
Steatosis stage	S0	13 (21.3)	16 (26.2)	0.891
	S1	15 (24.6)	15 (24.6)
	S2	18 (29.5)	15 (24.6)
	S3	15 (24.6)	15 (24.6)
NFS risk category	Low	9 (14.3)	9 (14.3)	0.852
	Intermediate	42 (66.7)	40 (63.5)
	High	12 (19.0)	14 (22.2)
FIB-4 risk category	Low	53 (79.1)	52 (77.6)	0.812
	Intermediate	11 (16.4)	11 (16.4)
	High	3 (4.5)	4 (6.0)

¹McNemar-Bowker test.

²Marginal homogeneity test.

FIB-4: Fibrosis-4 index; NFS: Nonalcoholic fatty liver disease fibrosis score.

Associations between biochemical parameters, lipogenesis biomarkers and imaging findings after 6 months of the SGLT2i therapy

After 6 months of SGLT2i therapy, circulating MTTP concentrations were positively associated with total cholesterol (ρ = 0.301, P = 0.01) and low-density lipoprotein (LDL) (ρ = 0.284, P = 0.02), whereas circulating PPAR-α concentrations were positively correlated with AST (ρ = 0.250, P = 0.04), as shown in Table 6. Furthermore, UGAP correlated with triglycerides (P < 0.001), and both 2D-SWE and UGAP values were positively associated with markers of liver injury, including ALT, AST, and GGT, while showing inverse associations with high density lipoprotein (HDL) (P ≤ 0.04 for 2D-SWE and P ≤ 0.01 for UGAP). The strongest associations for both imaging modalities were observed with anthropometric measures, particularly waist circumference and BMI (P < 0.001 for all) (Table 7).

Table 6 Spearman correlation between laboratory parameters and lipogenesis-related biomarker concentrations after 6 months of sodium-glucose cotransporter 2 inhibitor therapy.

Parameter	PPAR-γ	PPAR-α	SREBP-1c	MTTP
Cholesterol (mmol/L)	0.109 (0.38)	0.026 (0.84)	0.037 (0.77)	0.301 (0.01)a
HDL (mmol/L)	0.001 (0.99)	0.125 (0.32)	-0.097 (0.44)	0.049 (0.70)
LDL (mmol/L)	0.065 (0.60)	-0.034 (0.79)	0.075 (0.55)	0.284 (0.02)a
Triglycerides (mmol/L)	0.081 (0.52)	0.069 (0.58)	0.199 (0.11)	0.075 (0.55)
ALT (U/L)	-0.097 (0.43)	0.222 (0.07)	0.210 (0.09)	0.089 (0.48)
AST (U/L)	-0.205 (0.10)	0.250 (0.04)a	0.069 (0.58)	0.007 (0.95)
ALP (U/L)	-0.092 (0.46)	0.082 (0.52)	-0.073 (0.56)	0.021 (0.87)
AST/ALT ratio	-0.032 (0.80)	-0.093 (0.45)	-0.274 (0.02)a	-0.141 (0.25)
GGT (U/L)	0.041 (0.74)	0.074 (0.56)	-0.090 (0.47)	0.085 (0.50)
Albumin (g/L)	0.028 (0.82)	-0.023 (0.86)	0.279 (0.02)a	-0.069 (0.58)
Prothrombin time INR	-0.144 (0.26)	-0.256 (0.04)a	-0.157 (0.22)	-0.177 (0.17)
Total bilirubin (µmol/L)	0.072 (0.56)	-0.105 (0.40)	0.004 (0.98)	-0.198 (0.11)
Conjugated bilirubin (µmol/L)	0.009 (0.95)	-0.025 (0.84)	-0.013 (0.92)	-0.300 (0.01)a

^aP < 0.05.

Values are Spearman’s correlation coefficients (ρ) with corresponding P values in parentheses. Concentrations of lipogenesis-related biomarkers were measured in ng/L. ALT: Alanine aminotransferase; ALP: Alkaline phosphatase; AST: Aspartate aminotransferase; GGT: Gamma-glutamyl transferase; HDL: High-density lipoprotein; INR: International normalized ratio; MTTP: Microsomal triglyceride transfer protein; PPAR-γ: Peroxisome proliferator-activated receptor gamma; PPAR-α: Peroxisome proliferator-activated receptor alpha; SREBP-1c: Sterol regulatory element-binding protein 1c.

Table 7 Spearman correlations of clinical, laboratory, and lipogenesis-related biomarkers with liver stiffness and hepatic steatosis (ultrasound-guided attenuation parameter) at baseline and after 6 months of sodium-glucose cotransporter 2 inhibitor therapy.

Parameter	Baseline		6 months
	2D-SWE	UGAP	2D-SWE	UGAP
Age (year)	0.079 (0.53)	-0.288 (0.02)a	-0.028 (0.82)	-0.254 (0.04)a
BMI (kg/m2)	0.563 (< 0.001)a	0.467 (< 0.001)a	0.467 (< 0.001)a	0.538 (< 0.001)a
WC (cm)	0.582 (< 0.001)a	0.467 (< 0.001)a	0.652 (< 0.001)a	0.551 (< 0.001)a
ALT (U/L)	0.172 (0.16)	0.458 (< 0.001)a	0.326 (0.01)a	0.465 (< 0.001)a
AST (U/L)	0.138 (0.27)	0.379 (< 0.001)a	0.258 (0.03)a	0.311 (0.01)a
ALP (U/L)	-0.072 (0.56)	-0.012 (0.93)	0.079 (0.53)	0.001 (0.99)
AST/ALT ratio	-0.104 (0.40)	-0.324 (0.01)a	-0.215 (0.08)	-0.423 (< 0.001)a
GGT (U/L)	0.167 (0.18)	0.513 (< 0.001)a	0.285 (0.02)a	0.517 (< 0.001)a
Cholesterol (mmol/L)	-0.125 (0.31)	0.110 (0.40)	-0.249 (0.04)a	-0.085 (0.50)
HDL (mmol/L)	-0.125 (0.31)	-0.132 (0.31)	-0.250 (0.04)a	-0.323 (0.01)a
LDL (mmol/L)	-0.063 (0.61)	0.048 (0.71)	-0.249 (0.05)	-0.05 (0.69)
Triglycerides (mmol/L)	0.057 (0.65)	0.369 (< 0.001)a	0.161 (0.20)	0.371 (< 0.001)a
UGAP	0.357 (0.005)a	-	0.422 (< 0.001)a	-
PPAR-γ	-0.068 (0.58)	-0.047 (0.72)	-0.233 (0.06)	0.045 (0.72)
PPAR-α	-0.016 (0.90)	-0.183 (0.16)	0.023 (0.85)	0.143 (0.25)
SREBP-1c	-0.080 (0.52)	-0.037 (0.78)	-0.127 (0.31)	0.050 (0.69)
MTTP	0.055 (0.66)	0.017 (0.89)	-0.268 (0.03)a	0.066 (0.60)
NFS	0.334 (0.01)a	0.050 (0.71)	0.236 (0.06)	-0.032 (0.80)
FIB-4	0.057 (0.65)	-0.141 (0.28)	0.014 (0.91)	-0.201 (0.10)

^aP < 0.05.

Values are Spearman’s correlation coefficients (ρ) with corresponding P-values in parentheses. Concentrations of lipogenesis-related biomarkers were measured in ng/L, two-dimensional shear wave elastography in kPa; ultrasound-guided attenuation parameter in dB/cm/MHz. 2D-SWE: Two-dimensional shear wave elastography; ALT: Alanine aminotransferase; ALP: Alkaline phosphatase; AST: Aspartate aminotransferase; BMI: Body mass index; FIB-4: Fibrosis-4 index; GGT: Gamma-glutamyl transferase; HDL: High-density lipoprotein; LDL: Low-density lipoprotein; MTTP: Microsomal triglyceride transfer protein; NFS: Non-alcoholic fatty liver fibrosis score; PPAR-γ: Peroxisome proliferator-activated receptor gamma; PPAR-α: Peroxisome proliferator-activated receptor alpha; SREBP-1c: Sterol regulatory element-binding protein 1c; UGAP: Ultrasound-guided attenuation parameter; WC: Waist circumference.

Liver stiffness assessed by 2D-SWE exhibited a significant positive correlation with hepatic steatosis measured by UGAP in the overall cohort (ρ = 0.422, P < 0.001). Additionally, 2D-SWE demonstrated a significant negative correlation with the biomarker MTTP in the overall cohort (ρ = -0.268, P = 0.03) (Table 7). After therapy, ALT, mean corpuscular volume (MCV) and total protein differed between the NFS low-risk group and both intermediate- and high-risk groups (P = 0.01, P = 0.04, and P = 0.02, respectively), while the AST/ALT ratio and platelet counts differed across all three groups (P < 0.001), showing a progressive increase and decrease, respectively, with higher fibrosis risk. Erythrocyte count differed between NFS low- and high-risk groups (P = 0.01). Albumin levels differed between the NFS high-risk group and both low- and intermediate-risk groups (P = 0.04) (Table 8). Similarly, across FIB4 risk categories, the AST/ALT ratio increased and platelet and leukocyte counts decreased progressively from low- to high-risk groups, with all differences reaching statistical significance (P ≤ 0.01) (Table 8).

Table 8 Comparison of selected laboratory parameters across advanced fibrosis risk categories defined by the nonalcoholic fatty liver disease fibrosis score and Fibrosis-4 index after 6 months of sodium-glucose cotransporter 2 inhibitor therapy.

Parameter	Low risk	Intermediate risk	High risk	P value1
NFS	n = 9	n = 41	n = 15
ALT (U/L)	38 (33.5-84)	35 (29.5-59)	28 (19.5-35.5)	0.01a
AST (U/L)	34 (24.5-54)	26 (23.5-40.5)	23 (18-28)	0.18
AST/ALT ratio	0.7 (0.6-0.9)	0.7 (0.7-0.9)	0.9 (0.7-1.1)	< 0.001b
Erythrocytes	4.8 (4.6-5.3)	5.3 (4.9-5.8)	4.9 (4.4-5.1)	0.01c
MCV (fL)	84.5 (82.6-92.7)	84.7 (81.7-87.4)	88 (85.8-92)	0.04a
Platelet count (109/L)	282 (264.5-355.5)	318 (277.5-391)	240 (207-260.5)	< 0.001b
Leukocytes (109/L)	9.4 (7.4-10)	9.2 (6.9-10.6)	7.8 (6.6-9.1)	0.20
Albumin (g/L)	44.7 (44-46.3)	45.3 (44.1-48.7)	45 (43.4-47)	0.04d
Total protein (g/L)	74 (70-77.5)	77 (74-78)	72 (69-74.8)	0.02a
FIB-4	n = 52	n = 11	n = 4
ALT (U/L)	26 (20.3-35)	21 (17-29)	22.5 (15-38.3)	0.44
AST (U/L)	23.5 (20-28)	23 (21-31)	34 (22-44.5)	0.30
AST/ALT ratio	0.9 (0.7-1.1)	1.2 (0.7-1.4)	1.3 (1.2-1.7)	0.004a
Erythrocytes	5 (4.6-5.4)	5 (4.7-5.5)	4.3 (4.1-5)	0.15
MCV (fL)	88 (84.7-91.3)	88 (85.9-91.5)	88 (85.6-88.9)	0.90
Platelet count (109/L)	248 (214-283)	193 (174-221)	145.5 (139.3-177.3)	< 0.001a
Leukocytes (109/L)	8.3 (7.3-9.5)	7.6 (6.9-9.8)	5.5 (4.7-6)	0.01d
Albumin (g/L)	45.3 (43.1-47)	47.4 (43.1-47.7)	44.4 (41.3-46.2)	0.47
Total protein (g/L)	74 (70.4-77)	76 (71-77)	70 (70-72.3)	0.24

¹Kruskal-Wallis test (post hoc Conover test).

^aP < 0.05 were observed between the low-risk group and the intermediate- and high-risk groups.

^bP < 0.05 were observed between all three risk groups.

^cP < 0.05 were observed between the low- and high-risk groups.

^dP < 0.05 were observed between the high-risk group and the low- and intermediate-risk groups.

n denotes the number of patients in each fibrosis risk category. Values are presented as median (interquartile range). ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; MCV: Mean corpuscular volume; NFS: Nonalcoholic fatty liver disease fibrosis score; FIB-4: Fibrosis-4 index.

DISCUSSION

To our knowledge, this is among the first prospective studies to integrate the assessment of circulating lipogenesis biomarkers, quantitative ultrasound-derived liver parameters (2D-SWE and UGAP), and comprehensive laboratory and anthropometric measures following SGLT2i therapy in patients with T2DM. We hypothesized that SGLT2i therapy is associated with changes in circulating lipogenesis-related biomarkers and liver stiffness. Accordingly, we measured serum SREBP-1c, PPAR-α, PPAR-γ, and MTTP concentrations and assessed liver stiffness by 2D-SWE and hepatic steatosis by UGAP after six months of SGLT2i therapy. Our findings suggest that SGLT2i therapy is associated with modulation of selected circulating lipogenesis biomarkers and a modest reduction in liver stiffness measured by 2D-SWE whose clinical relevance remains uncertain. These findings may be consistent with early metabolic adaptations rather than definitive improvement in liver fibrosis.

Our cohort reflects a real-world population of patients with early-stage T2DM, prevalent obesity, and cardiometabolic comorbidities, representing the population at highest risk for MASLD owing to the close pathophysiological interplay between diabetes, obesity, and hepatic steatosis. As expected in such a population, many participants were receiving concomitant antihypertensive, lipid-lowering, glucose-lowering, and cardiovascular medications, which may have influenced hepatic lipid metabolism, inflammatory pathways and aminotransferase levels independently of SGLT2i therapy. Accordingly, the observed changes should be interpreted in the context of comprehensive cardiometabolic management rather than as being attributable exclusively to SGLT2i therapy. In addition, changes in lifestyle habits may also influence disease management, making it more difficult to disentangle changes associated with SGLT2i therapy from those related to other factors. In our cohort, however, no significant changes were observed in alcohol consumption, smoking status, physical activity, meal frequency, or dietary supplement use during follow-up, reducing the likelihood that major lifestyle changes contributed to the observed findings, although residual confounding cannot be excluded. Nevertheless, because SGLT2i are widely used in the treatment of T2DM, particularly among patients with coexisting heart failure or chronic kidney disease, many patients with MASLD are likely to receive these medications and may also derive benefit with respect to MASLD, potentially through weight loss and favorable hepatic metabolic effects[32]. However, the evidence supporting this benefit remains less well established than that for incretin-based therapies, particularly GLP-1 receptor agonists, and metabolic/bariatric surgery[33,34]. Therefore, SGLT2i should not be considered a substitute for therapies with stronger evidence for the treatment of MASLD. Rather, any potential benefit for MASLD should be viewed as an additional benefit for patients receiving these medications for their established indications. Six months of SGLT2i therapy was associated with significant improvements in glycemic control, accompanied by reductions in BMI and waist circumference, reflecting an overall improvement in metabolic status. These changes are consistent with the established systemic metabolic effects of SGLT2i and may contribute to mitigation of hepatic metabolic burden[35,36]. Furthermore, erythrocyte count, hemoglobin, and hematocrit were elevated following SGLT2i therapy, consistent with the established hematologic effects of this drug class. These alterations have been linked to favorable cardiovascular and renal outcomes in T2DM patients and are likely mediated by enhanced erythropoiesis[37]. Concomitantly, markers of hepatic synthetic function exhibited favorable modulation, reflected by increases in serum total protein and total bilirubin, along with marginal changes in coagulation parameters, collectively suggesting preserved or modestly improved hepatocellular functional capacity[38,39]. Stable liver enzymes and lipid profiles suggest that SGLT2i therapy may preserve hepatocellular integrity and metabolic control, consistent with prior evidence of potential neutral effects in T2DM[40]. Nevertheless, the absence of significant changes in ALT levels supports a cautious interpretation of our findings and suggests no measurable biochemical improvement in liver injury during follow-up.

Specifically, we observed a decrease in serum PPAR-γ concentrations and an increase in MTTP levels. PPARs are lipid-activated nuclear receptor transcription factors linked to metabolic disorders and comprise three isoforms: PPAR-α, predominantly expressed in the liver, PPAR-γ, mainly found in adipose tissue with lower hepatic expression, and PPAR-δ, primarily expressed in skeletal muscle with lesser expression in skin and adipose tissue[41]. In MASLD, hepatic PPAR-γ is upregulated, promoting fatty acid storage and lipid droplet formation. Preclinical studies have shown that SGLT2i therapy, including dapagliflozin, empagliflozin, and canagliflozin, reduces hepatic PPAR-γ expression, whereas in our study a decrease was observed in circulating PPAR-γ concentrations[14,19]. MTTP plays a critical role in the assembly and secretion of apolipoprotein B-containing lipoproteins, facilitating hepatic lipid export. MTTP deficiency impairs very LDL formation, leading to hepatic lipid accumulation and contributing to MASLD pathogenesis. In preclinical models of MASH, obesity, and T2DM, SGLT2i therapy (e.g., ipragliflozin) improved insulin sensitivity, increased MTTP expression, and attenuated hepatic lipotoxicity[17,42]. Similarly, circulating MTTP concentrations increased following SGLT2i therapy in the study cohort. However, because PPAR-γ and MTTP were measured in serum rather than liver tissue, their circulating concentrations may be influenced by multiple tissues, differences in secretion or clearance, systemic metabolic status, and assay-related factors. Consequently, they cannot be directly compared with changes in hepatic expression reported in experimental studies. By contrast, circulating PPAR-α, which regulates fatty acid β-oxidation, and SREBP-1c, a key driver of lipogenesis, did not change significantly, possibly because these transcription factors are less sensitive to short-term interventions and may only respond under specific metabolic or physiological conditions[14]. Together with the observed changes in circulating PPAR-γ and MTTP, these findings suggest that SGLT2i therapy may influence lipid metabolism–related pathways. However, because only serum biomarkers were measured, no conclusions can be drawn regarding the underlying hepatic molecular mechanisms.

Chronic metabolic stress driven by excess free fatty acids promotes lipotoxicity and oxidative stress, leading to hepatic stellate cell activation and extracellular matrix deposition, thereby increasing stiffness[3]. Although the reduction in liver stiffness reached statistical significance in our cohort, the absolute difference (6.2 kPa to 6.1 kPa) was clinically uncertain and may fall within the expected measurement error margin of 2D-SWE. Together with the minimal changes observed in hepatic steatosis, this finding may reflect early hepatic metabolic adaptations rather than established structural improvement or fibrosis regression. Fibrosis and steatosis stages remained largely stable, with a numerical increase in F0/F1 fibrosis, a decrease in F4 fibrosis, and near-equal distribution across steatosis stages. Consistently, most participants were classified as low risk by FIB-4, while the higher proportion of intermediate-risk NFS likely reflects metabolic factors, including BMI and diabetes status, rather than true fibrotic severity. NFS and FIB-4 risk categories remained largely stable, with non-significant and minor increases in high-risk participants likely reflecting metabolic factors or the limited accuracy of these scores for detecting ≥ F3 fibrosis[11]. The observed decrease in liver stiffness on 2D-SWE, together with stable fibrosis and steatosis stages and low-to-intermediate FIB-4 and NFS risk classifications, suggests early changes in liver tissue biomechanics that may precede measurable improvements in fibrosis or steatosis.

Circulating MTTP concentrations were associated with total cholesterol and LDL levels, consistent with the established role of MTTP in lipoprotein assembly and hepatic lipid export[43,44]. In addition, circulating PPAR-α concentrations showed a week positive association with AST, which may reflect metabolic adaptations associated with SGLT2i therapy[45,46]. As observed in previous studies, both 2DSWE and UGAP values correlate positively with liver transferases (ALT, AST, GGT), indicating that higher stiffness and steatosis reflect hepatocellular injury. UGAP values are also positively associated with triglycerides, and both UGAP and 2D-SWE are associated with adiposity measures (BMI, waist circumference) and inversely associated with HDL, highlighting their performance as noninvasive indicators of metabolic dysfunction and steatosis severity[47-49]. The negative correlation between UGAP and age further suggests age-related alterations in liver fat content and tissue properties that influence attenuation measurements[48]. To our knowledge, this is the first study to demonstrate a positive correlation between liver stiffness assessed by 2D-SWE and hepatic steatosis quantified by UGAP in patients with T2DM, supporting the complementary use of these quantitative ultrasound techniques for the non-invasive evaluation of liver involvement in individuals at increased risk of MASLD. Moreover, we identified a significant inverse correlation between circulating MTTP and 2DSWE, suggesting that impaired lipoprotein assembly may contribute to increased hepatic stiffness. Increases in the AST/ALT ratio and decreases in platelet count in higher NFS and FIB-4 categories, along with reductions in albumin and ALT across NFS risk categories, indicate higher fibrosis risk and early impairment of hepatic synthetic function, consistent with their inclusion in the calculation formulas[10,50]. Concurrent reductions in total protein, erythrocyte count, and MCV in higher NFS categories, and in leukocyte count in higher FIB-4 categories, likely reflect hepatocellular injury, systemic metabolic and inflammatory alterations, and progressive fibrotic remodeling in MASLD[51,52].

Although this study demonstrated significant modulation of circulating lipogenesis biomarkers and liver stiffness following SGLT2i therapy, the findings should be interpreted in light of several limitations. First, the prospective observational single-arm design without randomization precludes causal inference and limits our ability to determine whether the observed changes can be attributed solely to SGLT2i therapy. In addition, in the absence of a control group, the potential influence of the natural course of the underlying conditions and regression to the mean cannot be excluded. Second, the relatively small sample size may have limited statistical power and the generalizability of the findings. The study population comprised only patients with T2DM receiving SGLT2i therapy, limiting extrapolation to non-diabetic populations, and the six-month follow-up may have been insufficient to detect long-term changes in liver remodeling. Furthermore, although significant changes were observed in circulating PPAR-γ and MTTP concentrations, all investigated biomarkers were measured in serum rather than liver tissue. Consequently, changes in their circulating concentrations do not necessarily reflect changes in intrahepatic gene or protein expression and therefore cannot be interpreted as direct evidence of altered hepatic pathway activity, nor should these serum biomarkers be considered validated markers of MASLD severity or treatment response. Future studies integrating circulating biomarker measurements with liver tissue analyses are needed to better define their relevance in MASLD. Most participants had little or no fibrosis (F0/F1) at baseline, limiting the ability to detect clinically meaningful fibrosis regression during follow-up. Accordingly, although liver stiffness decreased significantly, the absolute reduction was modest and remained within the same fibrosis stage. As repeated same-visit measurements were not performed, it cannot be determined whether this difference exceeded the expected technical and biological variability of 2D-SWE, precluding definitive interpretation of its clinical importance. In addition, only dapagliflozin and empagliflozin were evaluated, limiting generalizability to other SGLT2i. The study-specific questionnaire was not designed to provide sufficiently detailed or validated quantitative assessment of lifestyle exposures to permit reliable multivariable adjustment for their potential effects. Therefore, we did not perform an exploratory post hoc multivariable analysis, as this could have created a misleading impression of adequately controlling for confounding. Although lifestyle-related variables, including physical activity and dietary habits, remained comparable over time, suggesting that the observed findings were unlikely to be explained by measurable lifestyle modifications, smaller changes in lifestyle, concomitant medications, or treatment adherence may not have been fully captured and could have contributed to the observed findings. More broadly, the high prevalence of concomitant antihypertensive, glucose-lowering, lipid-lowering therapies, especially statins, and cardiovascular therapies represents an additional potential source of confounding. Because the study was not designed or powered to evaluate medication subgroups, and detailed longitudinal analyses according to medication type and dose were not prespecified, exploratory post hoc subgroup analyses were not performed as they could have produced unstable estimates. Consequently, the independent contribution of SGLT2i therapy to the observed changes could not be fully distinguished from that of concomitant treatment. This limitation also applies to metformin, although participants prescribed fixed-dose SGLT2i/metformin combinations were already receiving stable metformin therapy before enrollment, and metformin treatment remained unchanged throughout the study. Furthermore, although weight loss is an expected effect of SGLT2i therapy, it is itself associated with improvements in hepatic steatosis and metabolic function, making it difficult to distinguish weight loss-mediated effects from those directly attributable to SGLT2 inhibition. Finally, liver assessment relied on non-invasive imaging and circulating biomarkers rather than histological evaluation, which remains the reference standard for assessing liver histology.

CONCLUSION

This study provides a comprehensive prospective evaluation integrating circulating lipogenesis-related biomarkers, quantitative ultrasound-derived liver stiffness (2D-SWE), hepatic steatosis (UGAP), and clinical, laboratory, and anthropometric parameters to characterize the metabolic and liver-related changes associated with SGLT2i therapy in patients with T2DM. In addition to the well-established metabolic benefits of SGLT2i, including improvements in BMI, waist circumference, fasting plasma glucose, and HbA1c, treatment was associated with changes in selected lipogenesis-related biomarkers, characterized by decreased circulating PPAR-γ and increased MTTP levels. Importantly, these changes in circulating lipogenesis-related biomarkers should not be interpreted as direct indicators of intrahepatic gene or protein expression and require validation against established hepatic outcomes, ideally including imaging and tissue-based measures. Liver stiffness measured by 2D-SWE showed a statistically significant reduction, but the absolute change was small and is not clinically meaningful. Furthermore, no significant changes in hepatic steatosis or ALT levels were observed, supporting a cautious interpretation of the liver-related effects of SGLT2i therapy. Liver stiffness correlated with UGAP-assessed hepatic steatosis and with MTTP, highlighting associations between molecular biomarkers, imaging findings, and metabolic function. Fibrosis and steatosis stages, as well as NFS and FIB-4 categories, remained largely stable, although numerical trends toward lower advanced fibrosis were observed. Collectively, these findings suggest that the early effects of SGLT2i therapy are predominantly metabolic, with no evidence of clinically meaningful structural improvements in the liver over six months. Therefore, any potential beneficial effects of SGLT2i in MASLD should be considered complementary to established therapies with demonstrated efficacy rather than a replacement for them. Whether the early metabolic, biomarker, and imaging findings observed in our cohort translate into clinically relevant improvements in hepatic steatosis and fibrosis remains to be established in larger studies with longer follow-up and validated hepatic outcomes, including histological endpoints.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge Ms. Blaženka Dobrošević, Specialist in Clinical Biochemistry, Department of Hematological Diagnostics and Coagulation, Clinical Hospital Center Osijek, Osijek, Croatia, and Dr. Albina Dumić, Specialist in Family Medicine, Osijek, Croatia, for their valuable expert advice.