Selenoprotein P deficiency in MASLD: association with insulin resistance and liver fibrosis: a prospective case-control study

Abstract

Background & aims

Metabolic dysfunction-associated steatotic liver disease (MASLD) is a globally escalating health concern. Selenoprotein P (SEPP1) is a hepatokine involved in selenium transport and antioxidant defense, with conflicting data on its role in MASLD. This study investigated serum SEPP1 as a potential non-invasive biomarker for disease severity and fibrosis staging.

Methods

This prospective case-control study enrolled 160 Egyptian participants (80 MASLD, 80 healthy controls). MASLD patients were stratified by fibrosis severity using vibration-controlled transient elastography (VCTE): non-significant fibrosis (< 8 kPa, n = 40) and significant fibrosis (≥ 8 kPa, n = 40). Anthropometric, biochemical (including HOMA-IR, lipid profile, liver enzymes), and SEPP1 (ELISA) measurements were compared.

Results

SEPP1 levels were significantly lower in MASLD patients versus controls (p < 0.001), with the lowest levels in the significant fibrosis subgroup (p = 0.025 vs. non-significant fibrosis). SEPP1 correlated inversely with BMI (r=-0.23, p = 0.004), HOMA-IR (r=-0.25, p = 0.001), and fasting insulin (r=-0.23, p = 0.004). MASLD patients exhibited higher insulin resistance, dyslipidemia, and liver enzymes (all p < 0.001). Logistic regression identified BMI (OR = 1.4, 95% CI:1.3–1.6) and HOMA-IR (OR = 1.4, 95% CI:1.1–2.0) as independent MASLD predictors.

Conclusions

Reduced SEPP1 levels are strongly associated with MASLD severity and hepatic fibrosis. Its inverse correlation with insulin resistance and stepwise decrease with advancing fibrosis position SEPP1 as a promising simple biomarker for metabolic dysfunction and non-invasive fibrosis risk stratification. This is particularly relevant in high-burden populations like Egypt, where accessible tools are urgently needed to guide early intervention, and further studies should explore whether SEPP1 modulation or selenium supplementation could mitigate liver fibrosis progression.

Introduction

Metabolic dysfunction-associated steatotic liver disease (MASLD) affected community is expanding globally at a fast pace among all ages, along with the upgrowing rates of obesity and metabolic syndrome. Currently, 38% of all adults have MASLD [1]. Its prevalence has ethnoracial variation, with Egypt having one of the top-notch rates in Middle East and North Africa region with 45.0% estimated prevalence [2].

It is intimately linked to cardiovascular disease, type 2 diabetes, and cancers, primarily through shared mechanisms of insulin resistance (IR), chronic inflammation, and oxidative stress [3, 4]. IR initiates a vicious cycle of hepatic steatosis and systemic metabolic dysfunction, which fuels a pro-inflammatory state, contributing to endothelial dysfunction and accelerated atherosclerosis, thereby worsening IR. MASLD independently increases myocardial infarction and stroke risk by 1.5–2×, mediated by systemic release of inflammatory mediators and hepatokines that promote plaque vulnerability [4].

MASLD linked to multitude of other risk factors including dysregulated systemic metabolism, genetics, disturbed adipokines, mitochondrial dysfunction as well as dysbiosis & disrupted gut–hepatic axis, all of these might contribute to the initiation and advancement of disease [5]. It can progress over time causing hepatic inflammation (metabolic-associated steatohepatitis, MASH), fibrosis, and eventually cirrhosis or more seriously hepatocellular carcinoma even without cirrhosis, so coupling of early identification of cases & proper management can abort progression to more serious stages and improve overall liver health [6]. Despite the solid role of liver biopsy in histological confirmation of MASLD stages and diagnosis, it has many drawbacks limiting its practicality for population screening [7]. Thus the improvement of non-invasive diagnostic tests and searching for helpful markers for early& accurate staging of liver disease is now crucial as it enables timely detection of high-risk individuals, optimizes resource allocation, and facilitates targeted interventions, eventually improving outcomes via limiting disease burden on individuals, community as well as saving resources globally [8].

Selenium is an essential micronutrient that holds a particular significance for human life, being an indispensable component of Selenoproteins and some antioxidant enzymes such as glutathione peroxidase &iodothyronine deiodinase, it is also crucial for the integrity of the immune function through boosting macrophages &T cells activity as well as production of antibodies [9].

Human body carries twenty-five Selenoprotein genes involved mainly in thyroid hormone metabolism & oxidative stress regulation in various tissues, 8 of them encode glutathione peroxidases, three genes for thioredoxin reductases, three for Iodothyronine deiodinase, one gene for selenophosphate synthase-2. The remaining Selenoproteins include Sep15, SelH, SelI, SelK, SelM, SelN, SelO, SelP, SelR, SelS, SelT, SelV, and SelW [9].

Selenoprotein P (SEPP, SEPP1 or SelP) is an extracellular glycoprotein hepatokine serving as the main selenium (Se) transporter from liver to various body tissues & also a potent antioxidant & anti carcinogenic [10]. Approximately 60% of plasma Se is carried by SEPP1 via its structure that incorporates Selenium in as many as ten Selenocysteine residues compared to one residue in other Selenoproteins, it can be considered as a biomarker of Se status reflecting the bioavailable Se to target tissues [11] Se intake varies widely worldwide being lower in Europe than in the United States, while higher intake in the Amazonas region, due to variation in selenium availability in the soil, selenium type, different Se requirement among genders & absorptive mechanism. Selenium deficiency reduces the expression and activity of Selenoproteins, but the body prioritizes SEPP1 production to transport Se to vital extra-hepatic tissues like the brain, even at the cost of other Selenoproteins, Sever prolonged deficiency compromised overall Selenoproteins function. Genetic mutations affecting insertion of selenocysteine in Selenoproteins also disturb Selenoprotein level [12]. Studies from Middle East showed varying results even within areas from same country. Limited data about Se intake in Egypt, however the available studies from Egypt reported serum Se of less than 90 µg/L in healthy control subjects. According to Global estimation of dietary micronutrient, inadequate intakes were common in countries in South Asia and Africa, including Egypt [13, 14].

SEPP1 is considered one of the housekeeping proteins which are essential for cellular maintenance & basic function. Stability of its plasma concentration is seen under physiological conditions along with stable Selenium intake, but significant alteration in its level was observed under pathophysiological conditions, as malignancies, pulmonary arterial hypertension & appeared to be closely tied to disturbed lipid and glucose metabolism as in metabolic syndrome, type II diabetes mellitus (T2DM) as well as MASLD [15, 16].

Notably, other hepatokines like osteopontin (encoded by SPP1) have been directly implicated in cardiovascular risk; elevated serum osteopontin (SPP1) levels confer an increased risk for plaque vulnerability in patients with coronary artery disease, illustrating the critical role of liver-derived factors in systemic cardiometabolic pathways [17].

The evidence surrounding SEPP1 disturbance remains inconclusive, with results varying between high SEPP levels in disturbed glucose metabolism & MASLD & quite opposite results [18–20].

This high regional MASLD burden necessitates accessible biomarkers for local use. Since global selenium intake varies widely and directly influences SEPP1 levels, population-specific studies are crucial to validate its biomarker utility.

Objectives

This study aimed to evaluate the serum level of SEPP1 in patients with MASLD compared to healthy participants & to detect its correlation with disease severity, fibrosis staging and metabolic parameters.

Patients and methods

Study design

This is a prospective case control study conducted on participants attending Kasr Al Ainy internal medicine outpatient clinic in the period from May 2024 to February 2025.

Ethical considerations

Before enrolment in this study, informed written consent was obtained from all participants after explaining the purpose of the study and assuring confidentiality. The study adhered to the ethical principles of Revised Helsinki Declaration & was accepted by the Research Ethical Committee of the Faculty of Medicine-Cairo University (MS-298-2023).

Participants

A consecutive convenience sampling method was used. All eligible patients attending the Kasr Al Ainy internal medicine and hepatology outpatient clinic during the study period (May 2024 to February 2025) who met the inclusion and exclusion criteria were sequentially invited to participate until the target sample size was reached, 80 participants diagnosed as MASLD and 80 healthy participants (sex & age matched) as a control group were all healthy volunteer. The control group (group 1): has a normal liver texture by ultrasonography, while MASLD group further subdivided based on results of Vibration-Controlled Transient Elastography (VCTE) & Controlled Attenuation Parameter (CAP) into: (group 2): 40 patients with any severity of liver steatosis and fibrosis ˂ 8 kPa (non-significant fibrosis), while the other 40 patients (group 3) had any severity of liver steatosis and fibrosis ≥ 8 kPa (significant fibrosis). All participants with MASLD were included after omitting individuals with acute or chronic liver disease due to viral infection, alcoholic hepatitis, hepatic malignancy, any chronic illness (like diabetes, hypertension, chronic renal & thyroid disease). Also, pregnant females were excluded from the study. All MASLD participants are included according to MASLD diagnostic EASL practice guidelines 2024 [21, 22].

Methods

All participants underwent detailed clinical evaluation via an in-depth medical history and clinical examination, including measurement of (body weight, height, waist circumference and body mass index, or “BMI”), no one of our participants had history of alcohol intake, detailed biochemical examination & assessment of the liver by ultrasound (US) and fibroscan to assess degree of hepatic steatosis and liver stiffness.

Data on potential confounders, including lifestyle and dietary data, including detailed physical activity levels, dietary patterns, selenium intake, assessment of other micronutrient statuses, and use of supplements or antioxidants, detailed family history of metabolic diseases (e.g., type 2 diabetes, dyslipidemia), were not systematically collected in this initial protocol.

Biochemical examination

Sampling

After overnight fasting, venous blood samples were withdrawn aseptically from every participant. Blood samples were partitioned into three equal aliquots for subsequent analysis: the first was collected into EDTA-containing tubes for determination of glycated heamoglobin (HbA1 C), the second was collected into fluoride-containing tubes for determination of fasting blood glucose (FBG) and the last part was collected into serum separating tubes for determination of lipid profile including {Triglycerides (TGs), total cholesterol, low density lipoprotein (LDL), High-density lipoprotein (HDL) &very low-density lipoprotein (VLDL), Serum Alanine aminotransferase (ALT), Aspartate aminotransferase (AST), gamma-glutamyl transferase (GGT), bilirubin, albumin as well as virology screening for hepatitis C and B (HCVAb and HBVsAg), Fasting insulin, and Selenoprotein P.

Serum preparation

Following blood collecting, the whole blood was allowed to clot by leaving it undisturbed at room temperature, a process that usually takes 20 min. The clot was then removed by centrifugation at 3,000 rpm for 20 min. Resulting serum was aliquoted into multiple aliquots and kept at -20 c until analyzed.

We assessed Insulin resistance via the homeostasis model of assessment (HOMA-IR) formula: [fasting blood glucose (mg/dL) × fasting insulin (IU/mL)] /405.

Regarding ELISA specifications for Selenoprotein P quantification: Sensitivity was 0.01 ng/ml, SEPP1concentrations in our study were measured using a commercially available ELISA kit (INNOVA BIOTECH, Cat. No. In-Hu3585). According to the manufacturer, the assay is based on a sandwich ELISA principle using antibodies specific to SEPP1, which ensures target specificity [23]. This design minimizes non-specific binding. The precision values are reported by manufacturer intra-assay precision (CV < 10%) and inter-assay precision (CV < 12%). Assay range by kit 0.1ng/ml -10 ng/ml. Samples were diluted 1:5 prior to analysis to accommodate higher sample concentration according to manufacturer’s instruction provided in the kit insert (Dilution factor X 5) So, the effective measuring range: 0.5–50 ng/ml. All samples were processed according to the kit protocol and standard curves were included in each run to ensure accuracy.

The optical density (OD) which is proportional to the concentration of SEPP1, was measured spectrophotometrically at a wavelength of 450 nm. SEPP1 concentration in the samples was quantified using a standard curve generated by plotting a known concentrations of SEPP1 and its corresponding OD reading on log-log scale (Fig. 1) & concentration of SEPP1 in a sample is determined by plotting the sample’s O.D. on the Y-axis & the corresponding SEPP1 concentration is read from the curve.

Fig. 1.

Human Selenoprotein P (SEPP1) standard curve

Assessment of hepatic steatosis by ultrasound

In our research all the participants were assessed by abdominal ultrasonography, using the same machine operator (MH) utilizing a high-resolution multifrequency B-mode scanner (SDD-5500; Aloka, Tokyo, Japan) 2.5-5.0 MHz transducer. Based on the ultrasonographic data regarding liver’s echogenicity (brightness) compared to the renal cortex, the clarity of intrahepatic blood vessels and diaphragm & the presence of deep beam attenuation, we classified MASLD into mild, moderate, and severe. severe steatosis when there is increased hepatic echogenicity along with vascular blurring and beam attenuation (Higher echogenicity of the liver causes insufficient beam penetration, resulting in compromised visualization of the posterior hepatic segment and the hepatic vessels) [24].

VCTE examination

Performed for all by a skilled radiologist (blind to the liver US results) using the FibroScan^® Mini + 430 (Echosens, Paris, France). Reliable procedures had an interquartile range/median (1QR/M) ratio of less than or equal to 30%, ten validated measures, and a success rate of at least 60%. Degree of fibrosis & liver stiffness measurement (LSM) assessed by fibroscan are expressed in kilopascals (kPa) & values < 8 kPa rule out significant fibrosis, between 8 and 12 kPa is F3 (significant fibrosis) and > 12 kPa is F4 (cirrhosis) [19]. The controlled attenuation parameter (CAP) test was employed to quantify the hepatic steatosis degree, and the results were expressed in decibels/meter (dB/m). CAP values for liver steatosis ranged from S0 (no steatosis) to S3 (severe steatosis). The CAP values for S1, S2, and S3 are 248 dB/m, 268 dB/m and 280 dB/m respectively [22].

Statistical methods

Data analysis was proceeded using statistical package of social science (SPSS Version 28; IBM Corp., Armonk, NY). Descriptive data were summarized using means, standard deviation (SD), median and ranges for numerical data & frequencies and percentages for categorical variables. Comparisons between 2 groups for numerical variables performed using the student t-test or the non-parametric Mann–Whitney U test, while Chi-square test or Fisher’s exact test were used for comparing categorical variables. For multiple group comparisons, the one-way analysis of variance (ANOVA) or Kruskal-Wallis test was performed, with post hoc analysis as appropriate. Spearman’s correlation coefficients applied for correlations between variables. To analyze the independent influence of various variables on MASLD risk, variables with a significance level of less than 0.10 in the univariate analysis were included in a stepwise logistic regression model. This regression analysis was conducted to calculate adjusted odds ratios (ORs) and quantify the magnitude of association between each risk factor and MASLD. odds ratios and 95% Confidence Interval (95% CI) were done also (95% CI that doesn’t contain 1.0 is considered significant). All statistical tests will be two-tailed, and a p-value ≤ 0.05 will be considered statistically significant.

Results

Our MASLD & healthy volunteers were sex and age matched, females constituted 50% of both our cases & control. The mean age of cases and controls was (35 ± 7 years & 32 ± 8 years), in order. Comparing the anthropometric & biochemical parameters of both MASLD & control groups (as elaborated in Tables 1 and 2), we observed significantly higher BMI and waist circumference in MASLD patients (p value < 0.001). Furthermore, the MASLD group demonstrated a statistically higher pattern of dyslipidemia (indicated by higher total cholesterol, LDL, triglycerides, VLDL & lower HDL) along with significantly higher level of liver enzymes (AST, ALT, GGT, ) HOMA-IR, fasting insulin, fasting blood sugar & HbA1c compared to healthy controls with (p value < 0.001) for all listed biochemical parameters. Significantly low serum levels of (Selenoprotein P& albumin were observed among the participants compared to the controls (p value < 0.001) as plotted on (Fig. 2). Comparable bilirubin levels were observed among both study groups, with no statistically significant disparity (p value 0.107).

Table 1.

Show the anthropometric measurements & biochemical parameters in MASLD patients and controls. In this table, data are presented as mean ± SD for normally distributed variables and median (range) for non-normally distributed variables. Show the normally distributed variables

Parameter	Control (n = 80)	MASLD (n = 80)	p value
	Mean ± SD	Mean ± SD
Weight (kg)	69.9 ± 10.8	91.0 ± 17.5	< 0.001
BMI (kg/m²)	24.9 ± 3.3	32.9 ± 6.1	< 0.001
Waist circumference (cm)	83.8 ± 9.1	101.5 ± 16.0	< 0.001
Fasting blood glucose (mg/dL)	85 ± 11	92 ± 14	< 0.001
HbA1c (%)	5.2 ± 0.4	5.5 ± 0.6	< 0.001
Total cholesterol (mg/dL)	188 ± 34	206 ± 44	0.005
HDL-cholesterol (mg/dL)	50 ± 11	45 ± 11	0.002
LDL-cholesterol (mg/dL)	118.4 ± 30.4	134.2 ± 39.2	0.005
Triglycerides (mg/dL)	93 ± 47	138 ± 66	< 0.001
VLDL-cholesterol (mg/dL)	18.7 ± 9.5	28.1 ± 13.9	< 0.001
Serum albumin (g/dL)	4.7 ± 0.4	4.5 ± 0.3	0.015

BMI body mass index, HbA1C glycated heamoglobin, HDL high-density lipoprotein, LDL low-density lipoprotein, TG triglycerides, VLDL very low-density lipoprotein, SD Standard deviation, p value < 0.05 is considered significant

Table 2.

Show the anthropometric measurements & biochemical parameters in MASLD patients and controls. In this table, data are presented as mean ± SD for normally distributed variables and median (range) for non-normally distributed variables. Non-normally distributed variables

Parameter	Control (n = 80)	MASLD (n = 80)	p value
	Median (range)	Median (range)
AST (U/L)	19 (10–103)	22 (8–114)	0.001
ALT (U/L)	14 (6–68)	25 (7–156)	< 0.001
GGT (U/L)	16 (8–137)	25 (7–192)	< 0.001
Total bilirubin (mg/dL)	0.5 (0.1–1.5)	0.4 (0.1–1.6)	0.107
Fasting insulin (µIU/mL)	5.9 (0.3–26.8)	10.7 (0.3–32.3)	< 0.001
HOMA-IR	1.2 (0.1–7.8)	2.4 (0.1–9.2)	< 0.001
Selenoprotein P (ng/mL)	8 (3–50)	5 (2–50)	< 0.001

AST aspartate aminotransferase, ALT alanine aminotransferase, GGT gamma-glutamyl transpeptidase, HOMA-IR Homeostatic Model Assessment for Insulin Resistance

Fig. 2.

Boxplot representing Selenoprotein level among MASLD & controls. Figure highlights the significantly low serum levels of SEPP1 among the MASLD participants compared to the controls

The independent influence of different variables on MASLD incidence was done using stepwise logistic regression as previously described. The analysis model identified BMI & HOMA-IR as significant predictors, with odds ratios (OR) of 1.4 (95% CI: 1.3–1.6) for BMI and 1.4 (95% CI: 1.1–2.0) for HOMA-IR, p value was (< 0.001 & 0.035) respectively.

Regression coefficient shows the effect of each variable after controlling other variables in the model. The model shows that BMI and HOMA-IR were the most crucial predictors of MASLD. For each unit increase in BMI & HOMA-IR there is a 40% increase in MASLD risk (displayed in Fig. 3).

Fig. 3.

Forest plot represents variables which are significant in logistic regression

Our results of VCTE further subgrouped MASLD participants into group 2: forty participants with any degree of liver steatosis and fibrosis ˂ 8 kPa (non-significant fibrosis) & group 3: forty participants with any degree of liver steatosis and significant fibrosis (8–12 kPa).Major bulk of our MASLD participants had no fibrosis (33.8%) & a minority (2.5%) had liver stiffness measurement > 12 kPa, while those with liver stiffness (8–12 kPa) represented by 47.5% of all involved MASLD participants in this study. Mean Kpa showed a statistically significant difference between group 2 & 3 being (5 ± 1.1) for group 2 & (8.7 ± 1.4) for group 3 with p value (< 0.001) respectively. Regarding the degree of steatosis, most of our MASLD participants were S3 (42.5%), followed by S2 (38.8%) & S1 represented by only (18.8%) of all MASLD enrolled participants, the mean CAP was 287 ± 45 for group 2 & 303 ± 42 for group 3 with no significant difference between both regarding steatosis p value was 0.111 (Table 3).

Table 3.

Comparing (mean ± SD) Kpa & CAP for both MASLD groups (group 2& group 3)

	Group 2	Group 3	P value
	Mean ± SD (range)	Mean ± SD (range)
KPa	5 ± 1.1 (3-6.6)	8.7 ± 1.4 (7-13.6)	< 0.001
CAP	287 ± 45 (224–378)	303 ± 42 (240–400)	0.111

SD Standard deviation, p value < 0.05 is considered significant

Comparison between the 3 groups (illustrated in Table 4) detect a significantly lower cardiometabolic risk factors (weight, BMI and waist circumference) in control compared to group 2&3 (p value for each is < 0.001) & on the contrary side these parameters showed nearly comparable levels among group 2&3 with no statistically significant variation.

Table 4.

Comparison between the three groups regarding clinical & laboratory data

	Group 1 (Control) (n = 80)	Group 2 (n = 40) non-significant fibrosis	Group 3 (n = 40) significant fibrosis
	Median (range)	Median (range)	Median (range)	p value	Pairwise comparison
Selenoprotein P (ng/dl)	8 (3–50)	7 (2–50)	5 (2–50)	< 0.001	***0.025, **<0.001, *0.752
AST	19 (10–103)	22 (8–80)	23 (12–114)	0.004	*0.090, **0.006, ***0.429
ALT	14 (6–68)	22 (8-156)	27 (7-130)	< 0.001	*<0.001, **<0.001, ***0.504
γGT	16 (8-137)	21 (8-134)	29 (7-192)	< 0.001	*0.012, **<0.001, ***0.393
Bilirubin	0.5 (0.1–1.5)	0.4 (0.1–0.95)	0.5 (0.2–1.6)	0.063
HOMA-IR	1.2 (0.1–7.8)	1.8 (0.1–6.6)	3 (0.7–9.2)	< 0.001	***0.008, **<0.001, *0.069
Fasting insulin	5.9 (0.3–26.8)	8.8 (0.3–31)	12.7 (3.8–32.3)	< 0.001	***0.017, **<0.001, *0.074
	Mean ± SD	Mean ± SD	Mean ± SD
Fasting blood sugar	85 ± 11	89 ± 14	95 ± 14	< 0.001	***0.049, **<0.001, *0.431
HbA1c (%)	5.2 ± 0.4	5.5 ± 0.7	5.5 ± 0.5	< 0.001	*<0.001, **<0.001, ***1
Albumin	4.7 ± 0.4	4.5 ± 0.3	4.5 ± 0.4	0.046	*0.067, **0.271, ***1
Total cholesterol	188 ± 34	199 ± 35	213 ± 52	0.006	*0.438, **0.004, ***0.371
HDL	50 ± 11	43 ± 9	46 ± 12	0.004	*0.004, **0.231, ***0.568
LDL	118.4 ± 30.4	128.6 ± 29.3	139.8 ± 46.7	0.007	**0.006, ***0.470, *0.397
TGs	93 ± 47	139 ± 61	138 ± 71	< 0.001	*<0.001, **<0.001, ***1
VLDL	18.74 ± 9.48	27.16 ± 12.57	28.95 ± 15.13	< 0.001	*<0.001, **<0.001, ***1
Weight (Kg)	69.9 ± 10.8	87.3 ± 18.9	94.8 ± 15.4	< 0.001	*<0.001, **<0.001, ***0.151
BMI (kg/m²)	24.9 ± 3.3	31.3 ± 6.5	34.4 ± 5.3	< 0.001	*<0.001, **<0.001, ***0.058
Waist circumference (Cm)	83.8 ± 9.1	99.5 ± 14.9	103.5 ± 17	< 0.001	*<0.001, **<0.001, ***0.601

Results in this table were organized based on the type of statistical measure used, with findings expressed as Mean ± SD presented separately from those reported as medians

p value < 0.05 is considered significant

SD Standard deviation

* =1 vs. 2

** =1 vs. 3

*** = 2 vs. 3

Lower Selenoprotein P level found in MASLD group with non-significant fibrosis compared to control but, with no real significance (p value 0.752), while in MASLD participant with significant fibrosis, Selenoprotein P showed its lowest level compared to control & MASLD with non-significant fibrosis p value < 0.001&0.025, respectively (as shown in Table 4 & Fig. 2).

Liver enzymes (AST, ALT & GGT) are significantly higher in MASLD with significant fibrosis followed by group 2 compared to group 1 with p values {G3 vs. G1 were (0.006, < 0.001 & <0.001)} & G2 vs. G1 (0.090, < 0.001, 0.012) consecutively, with no statistical significance between (G2 & G3).No statistical significance reported between all groups regarding serum albumin level.

Significantly higher readings of HOMA-IR, fasting insulin & fasting blood sugar were observed in group 3 versus group 1&2 with (p value < 0.001, 0.008), (< 0.001, 0.017). (< 0.001, 0.049 respectively). Despite their values being also higher in G2 compared to G1, but with no statistical significance, p values (0.069, 0.074 & 0.431) in the same order. (Shown in Fig. 4). Furthermore, group 2&3 exhibited a significantly higher HbA1c compared to control (p value < 0.001 for both).

Fig. 4.

Boxplot representing HOMA-IR level among the three study groups. Illustrates the significantly higher HOMA-IR values in group 3 (steatosis with significant fibrosis) compared to the 2 other groups

Regarding lipid profile

Control group showed significantly lower total cholesterol & LDL values compared to group 3 (p value 0.004 & 0.006) respectively, significantly higher HDL compared to group 2 (p value 0.004) & a significantly lower VLDL & triglycerides compared to both group 3 & 2 (p value < 0.001) for each in both groups. Values of lipid profile are nearly comparable between group 2 & 3.

We found that SEPP has a significant negative correlation with (BMI, fasting blood sugar, HOMA-IR and fasting insulin), as in Table 5.

Table 5.

Correlation between Selenoprotein with different factors

Factors	Selenoprotein
	r	p value	Degree of correlation
Age	-0.11	0.167	No significant correlation
Waist circumference	-0.11	0.175	No significant correlation
BMI	-0.23	0.004	Significant little negative correlation
Fasting blood sugar	-0.20	0.013	Significant little negative correlation
HOMA-IR	-0.25	0.001	Significant fair negative correlation
HbA1c	-0.15	0.059	No significant correlation
Fasting insulin	-0.23	0.004	Significant little negative correlation
Total cholesterol	-0.15	0.063	No significant correlation
HDL	-0.11	0.162	No significant correlation
LDL	-0.10	0.193	No significant correlation
Triglycerides	-0.09	0.280	No significant correlation
VLDL	-0.09	0.283	No significant correlation
KAP	-0.21	0.060	No significant correlation
CAP	-0.10	0.388	No significant correlation

r is the correlation coefficient (+ 1 denotes positive correlation, -1 denotes negative correlation) (r range from 0 to 0.25 (-0.25) =little or no correlation, From 0.25 to 0.50 (-0.25 to 0.50) =fair degree of correlation, From 0.50 to 0.75 (-0.50 to -0.75) =moderate to good correlation, Greater than 0.75 (or -0.75) =very good to excellent correlation.), p value < 0.05 considered significant

Discussion

MASLD is a confirmed globally underestimated nightmare, with its incidence progressing rapidly on the same marathon run side to side with the heightened incidence of obesity, insulin resistance, metabolic syndrome & type II diabetes mellitus, being now a master cause of chronic liver disease with its spectrum extending from simple steatosis up to cirrhosis or more fatal hazards as hepatocellular carcinoma [25].

Selenoprotein P is a potent antioxidant hepatokine serving mainly as a selenium conveyor to the various body target cells, although overproduction of serum SEPP being closely linked to hyperglycemia, insulin resistance, T2DM & vascular endothelial dysfunction through disturbing insulin signal transduction in the liver and skeletal muscle, others reported its inverse correlation with HOMA-IR & BMI. Newly, many conflicting threads connect Selenoproteins particularly SEPP to MASLD risk & development [20, 26, 27]. Considering this, the main purpose of this study & dedication of our efforts was to solve the previous conflict & to correlate serum SEPP level with progression & severity of MASLD, searching for a novel therapeutic pathway for the disturbed liver metabolic process in MASLD.

Overweight & obesity are potent accusers associated with three times & half increased likelihood of developing MASLD [28].

In our current study, significantly higher cardiometabolic risk factors (weight, BMI and waist circumference) were observed in MASLD patients (group 2&3) compared to control (p value for each is < 0.001). This is in agreements with results of Villarsi and Sungkar, Pirola et al. & Li L et al [28–30].

In this study, MASLD patients (G3 followed by G2) demonstrated significantly elevated AST, ALT & GGT compared to control with p values {G3 vs. G1 were (0.006, < 0.001 & <0.001)} & G2 vs. G1 (0.090, < 0.001, 0.012), which goes hand in hand with results found by Polyzos et al. & Choi et al., who reported significantly elevated AST, ALT in MAFLD versus control with (p values < 0.001) for each [20, 31].

Now, It is a well-established fact that MASLD is closely linked to Type 2 DM, IR & dyslipidemia, cardiovascular events by many pathophysiological mechanisms & MASLD patients have 1.5–2× higher risk of MI and stroke [32].

Substantial evidence links MASLD to early atherosclerosis, such as: Increased carotid intima-media thickness in obese patients. This is reinforced by serum eotaxin, a cytokine that recruits macrophages, which correlates with pro-inflammatory markers (TNF-α, IL-6) and independently predicts carotid thickening alongside steatosis severity. The pathophysiological connection between MASLD and cardiovascular disease is anchored in chronic inflammation and insulin resistance. These same drivers of hepatic steatosis and fibrosis also promote endothelial dysfunction, vascular smooth muscle proliferation, and unstable plaque formation, highlighting the need for biomarkers of these shared pathways [4, 33].

Further studies highlight the role of SEPP1 in promoting anti-inflammatory (M2-) macrophage polarization as observed by macrophage-specific deletion of SEPP1 impaired their ability to adopt a restorative (M2) phenotype in response to anti-inflammatory signals, leading to deficient repair and persistent inflammation in aged muscle niches [34].

This immunomodulatory role suggests that altered SEPP1 levels in MASLD could have implications beyond the liver, potentially affecting systemic inflammatory states and cardiovascular risk.

The pathophysiological connection between MASLD and cardiovascular disease is anchored in chronic inflammation and insulin resistance, often mediated by dysregulated hepatokines, SPP1 is a hepatokine which is directly implicated in increased plaque vulnerability in patients with coronary artery disease, in MASLD SPP1 is secreted from liver cells into the extracellular matrix, acting as a pro inflammatory cytokine mediating macrophage polarization toward pro-inflammatory phenotypes thus amplifying hepatic inflammation and oxidative stress. Activated macrophages further secrete SPP1, creating a positive feedback loop that sustains chronic inflammation, which promotes hepatocellular injury, fibrosis, and progression to more severe MASLD stages [17, 35]. This highlights a paradigm where liver-derived factors serve as molecular connectors between metabolic liver disease and its extra-hepatic complications. Recent bioinformatic analyses further identify SPP1, along with the inflammatory chemokines CXCL9 and IL2RB, as key shared biomarkers and therapeutic targets in the comorbidity of atherosclerosis and progressive MASLD (NASH), underscoring common inflammatory pathways [36] In this context, our finding of significantly reduced SEPP1 levels in advanced MASLD suggests it may be another crucial hepatokine in this network. While SPP1 is often elevated, SEPP1 appears depleted in advanced disease. This inverse pattern suggests a complex, stage-specific dysregulation of hepatokine signaling. The parallel reinforces the concept that the liver’s secretory profile is profoundly altered in MASLD, contributing not only to hepatic fibrosis but also to systemic metabolic dysfunction and potentially cardiovascular risk. The immunomodulatory role of SEPP1, particularly in macrophage polarization, further supports its potential involvement in these shared inflammatory pathways.

Our results deeply confirmed the strong bond between MASLD, Dyslipidemia & Chronic inflammation: Significantly higher values for total cholesterol & LDL in MASLD group with significant fibrosis compared to control (p value 0.004 & 0.006) respectively, a significantly lower HDL in MASLD group with non-significant fibrosis compared control (p value 0.004) & both MASLD groups showed significantly higher VLDL & triglycerides (TG) compared to control group with (p value < 0.001) for each. The MASLD group exhibited significantly higher adiposity, dyslipidemia, and insulin resistance compared with controls, all of which are established contributors to MASLD pathogenesis. These variables were therefore considered potential confounders and were adjusted for in multivariable analyses to assess the independent association between SEPP1 levels and MASLD (Fig. 3).

Ours support previous results spotted by Hegazy M et al. who reported a significant higher TG, total cholesterol, LDL-c & lower HDL among MAFLD patients vs. control [37].

Furthermore, similar results observed with Choi et al. who reported higher TG, total cholesterol & lower HDL-c among MAFLD compared to control & also, Çetindağlı et al. who detected significantly higher LDL-c and TG increased in MAFLD patients compared to controls [27, 31].

In addition, MASLD patients expressed significantly higher glycemic profile noted by higher HbA1c in both MASLD groups vs. control (p value < 0.001) as well as higher (HOMA-IR, fasting insulin & fasting blood sugar) especially noted in MASLD patients with significant fibrosis versus MASLD with non-significant fibrosis & control (p value < 0.001, 0.008), (< 0.001, 0.017). (< 0.001, 0.049 respectively).

In view of our data, BMI and HOMA-IR considered as the most significant predictors of MASLD with a linear correlation, where a single unit increase in BMI & HOMA-IR resulted in 40% escalation in MASLD risk. These results are consistent with the result of a previous cross-sectional study that enrolled non-diabetic adults in the US & concluded that HbA1c is an independent risk factor for NAFLD & correlated with steatosis severity, particularly in obese nondiabetic individuals [38]. Furthermore, this is consistent with earlier reports from previous Korean, Chinese & Egyptian studies with their opinion converge on the same conclusion that HbA1C & HOMA-IR are NAFLD independent risk & predictors non-diabetics [39–41].

SEPP is a selenium containing glycoprotein which is primarily produced by hepatocytes and secreted into circulation, it is a potent antioxidant [42] Understanding the clear correlation between SEPP & disturbed insulin signaling & glucose metabolism is still vague. Many studies reported significantly higher serum SEPP in insulin resistance, prediabetes & T2DM as well as in obese vs. lean, being positively correlated with cardiometabolic risks as visceral obesity, HOMA-IR, blood glucose & insulin level [18, 27, 31, 43, 44] The explanation for these results was based on the researches that revealed inhibitory effect of elevated SEPP levels on Adenosine monophosphate-activated protein kinase (AMPK) pathway, which can be reversed by use of SEPP-neutralizing antibodies or oral metformin use that activate AMPK which in turn suppress SEPP1 mRNA expression [45, 46].

Activation of AMPK pathway also negatively regulates de novo lipogenesis in the liver via the inhibitory phosphorylation of sterol regulatory element-binding protein 1c (SREBP1c) that is responsible for promoting the production of lipogenic enzymes, so suppression of AMPK will increase hepatic steatosis [47].

Obesity is characterized by a chronic low grade inflammatory condition with increased level of circulating pro inflammatory cytokines & higher load of oxidative stress & those factors upregulate SEPP as a stress response, leading to elevation of its serum levels, resulting in further suppression of AMPK & progression of IR.

In our study SEPP was significantly negatively correlated with BMI, fasting blood sugar, HOMA-IR and fasting insulin. This Align with a previous Egyptian study that reported inverse association between SEPP & markers of insulin resistance and obesity (glucose, insulin, HOMA-IR, BMI) & results from Polyzos et al. who reported significant negative correlation between SEPP and BMI [20, 48].

Our estimation of plasma SEPP1 level revealed a significantly lower level in MASLD participants compared to controls, with nadir values detected among the MASLD group with significant fibrosis (p < 0.001 vs. control; p = 0.025 vs. non-significant fibrosis), shown in Fig. 5. This finding of reduced SEPP1 in advanced MASLD contrasts with some literature reporting elevated levels in metabolic disorders. This discrepancy may stem from population-specific factors (e.g., genetic background, selenium status), assay variations, or, most notably, the disease stage of the cohort. Our results align with studies in advanced NASH and cirrhosis, suggesting SEPP1 dynamics may be phase-specific: potentially elevated in early metabolic stress but depleted with progressive hepatic injury, fibrosis, and antioxidant exhaustion. An analysis for extreme outliers (> 3 SD from the mean) confirmed this pattern was not driven by data anomalies.

Fig. 5.

Boxplot representing Selenoprotein P level among the three study groups. The figure depicted the significantly lower SEPP1 level in MASLD participants relative to controls, with the most pronounced reduction seen in MASLD group with significant fibrosis

Our explanation for this difference in results is regarding inverse correlation with BMI, HOMA-IR, fasting glucose

The increase in SEPP provoked by the stress of obesity induced inflammatory status might act as a protective antioxidant early, but in chronic settings, excess SEPP will impair insulin signaling by interfering with AMPK, reducing metabolic flexibility, further increasing hepatic lipogenesis & eventually leading to steatohepatitis & fibrosis that will end by reduced hepatic SEPP production due to advancing of liver pathology supporting the theory of dysregulated SEPP expression in advanced liver disease & this explain why the lowest SEPP level in our study observed in MASLD with significant fibrosis compared to those with non-significant fibrosis & control.

Also, long standing exhaustion of SEPP antioxidant pool will finally end by reduced SEPP due to consumption adding to this diminished capacity to combat oxidative stress in participant with MASLD & significant fibrosis. Tverezovska I et al. on studying NAFLD patients reported that Selenoprotein P inversely correlated with plasma level of pro-inflammatory cytokine (IL-8) and directly with the anti-inflammatory (IL-10), suggesting depletion of antioxidative SEPP due to worsened liver inflammation [43]. Another further explanation for the lowest SEPP level in MASLD with significant fibrosis in our results, the processes of long standing MASLD and associated metabolic disturbance may consume SEPP faster than the damaged liver can replenish and this may be aggravated in patients with selenium deficiency.

Earlier research enrolled German population supports our results, it reported that SEPP levels had a significant negative correlation with visceral & subcutaneous adiposity, metabolic syndrome & fasting blood glucose [49].

Our MASLD participants in G2 & G3 showed no significant steatosis difference & the mean of CAP (p value 0.111), but on the fibrosis side, there was a statistically significant difference between G2 & G3 in the mean of Kpa being significantly higher in G3 (p value < 0.001).The low SEPP level in both groups of MASLD participants vs. controls can be explained by presence of variable degrees of hepatic fibrosis in both group 2&3, that will lower SEPP level based on our previous explanation.

SEPP1 in MASLD is context sensitive, Yu R, et al. reported in his meta-analysis elevated SEPP in NAFLD participant who has other metabolic disorders with positive correlations to lipid and glucose metabolism markers [50].

Ours aligned with the results found in earlier study that noted a statistically significant difference between controls & all patients with hepatic steatosis{simple steatosis, borderline nonalcoholic Steatohepatitis (borderline NASH) & definite NASH} being highest in controls & lowest in definite NASH [48]. This finding is also concordant with other available literature that described lower SEPP1 levels detected in plasma samples from definite NASH candidates versus control [20].

No significant disparity regarding SEPP1 level between controls & MASLD group with non-significant fibrosis which also consistent with results from two studies reported non-significant variation between candidates with simple steatosis, borderline NASH & controls about SEPP level [20, 48]. This reinforces the theory of the more reduction in Selenoprotein p level, the more tendency for progression of hepatic fibrosis, owing this to imbalance between the excess of oxidative stress, lipotoxicity on hepatocytes and the body’s capacity to neutralize them [51, 52].

Effect of Se supplementation on MASLD patients was investigated in a pilot study performed in 2025 on 52 MASLD patients who were randomly grouped into 2 groups receiving either a food supplement containing choline, glutathione, selenium, zinc, and 2 probiotic strains once a day along with a balanced diet for 4 months or only balanced diet for 4 months, the intervention group showed significantly larger reductions in ALT, AST, and liver fat (CAP score), along with improvements in fibrosis markers, BMI, insulin resistance (HOMA‑IR), and lipid profiles compared to those receiving diet al.one [53]. Another previous study was in 2021 a study assessed effect of Selenoprotein gene variation & polymorphism; it was conducted on participants with a SEPP1 gene variant (rs7579 polymorphism) reported that SEPP1 variant rs7579 modulates how selenium affects lipid levels over time. Participants with this variant showed more long-term adverse changes in triglycerides, LDL, and total cholesterol when plasma selenium was high, suggesting that genetic variation influences selenium’s metabolic impact [54].

The correlation between Hepatic & circulating SEPP1: Hepatic SEPP1 serves as the Primary Source for serum SEPP1 as evidenced by a study that showed mice with selective deletion of Sepp1 in hepatocytes developed lowered plasma Sepp1 concentration and increased urinary selenium excretion, decreasing whole-body and tissue selenium concentrations [55]. MASLD can MASLD reduces both hepatic & circulating SEPP1: Bioinformatic analysis of liver tissue from NAFLD and NASH patients found decreased expression of multiple selenoprotein-related genes, indicating lower hepatic selenium handling in disease [56].

While this study demonstrates an independent association between SEPP1 and fibrosis severity, its potential clinical utility lies in its integration with established non-invasive tools. Future research should aim to develop and validate composite models that incorporate serum SEPP1 with scores such as the FIB-4 or NAFLD Fibrosis Score (NFS). Evaluating whether SEPP1 adds incremental diagnostic or prognostic value to these widely used algorithms could significantly enhance risk stratification in clinical practice, particularly for identifying high-risk patients within intermediate score ranges.

We should mention some Limitations of Our study, it was Cross-sectional; causality cannot be inferred, single-center design may limit generalizability, we did not collect detailed data on lifestyle factors such as dietary habits, physical activity, selenium intake, or the use of antioxidant supplements, these factors directly influence SEPP1 levels, insulin resistance, and liver enzymes, and their absence is an important limitation that may affect the interpretation of our associations and the generalizability of the findings, SEPP1 genetic variation wasn’t assessed and small sample size for advanced fibrosis subgroups, furthermore, while we discuss the established link between MASLD and cardiovascular disease, our study did not include direct vascular imaging (e.g., carotid Doppler ultrasound) to assess subclinical atherosclerosis in our participants. Future studies incorporating such measures are needed to directly examine the relationship between circulating SEPP1 levels and vascular health in the MASLD population, and a longitudinal study with dietary assessments are needed to validate SEPP1’s role in MASLD progression.

Conclusion

This study establishes that reduced serum Selenoprotein P (SEPP1) is associated with MASLD severity and significant hepatic fibrosis, showing an inverse correlation with BMI, HOMA-IR, and fasting insulin. This suggests SEPP1 depletion may reflect exhausted antioxidant capacity in progressive disease. Therefore, SEPP1 holds promise as a simple, scalable biomarker that could complement or triage the use of imaging modalities like VCTE. This is particularly relevant in resource-constrained or high-burden regions, where it could aid in prioritizing patients for further specialist assessment and intervention.

Recommendation

Based on our findings, we propose the following clinical and research pathways: Clinical integration: The combination of serum SEPP1 with non-invasive tools (e.g., VCTE) and metabolic markers (HOMA-IR, BMI) should be evaluated in prospective cohorts to develop integrated risk-stratification algorithms. Patients with low SEPP1 levels may be prioritized for intensive lifestyle intervention, evaluation of selenium status, and enhanced monitoring for fibrosis progression.

Therapeutic exploration

Further research is warranted to explore whether SEPP1 modulation or selenium supplementation can ameliorate insulin resistance and attenuate hepatic fibrosis in MASLD, representing a potential novel therapeutic avenue. And validation studies must define SEPP1’s additive value to scores like FIB-4 and establish population-specific cut-offs, while accounting for key confounders like selenium intake to confirm its clinical utility across diverse settings.

Abbreviations

MASLD

Metabolic dysfunction-associated steatotic liver disease

Se

Selenium

Sep or Sel

Selenoprotein

SEPP, SEPP1 or SelP

Selenoprotein P

SPP1

Secreted phosphoprotein 1

T2DM

Type II diabetes mellitus & Type 2 diabetes mellitus

VCTE

Vibration controlled Transient Elastography

CAP

Controlled Attenuation Parameter

IR

Insulin resistance

ORs

Odds ratios

CXCL9

C-X-C Motif Chemokine Ligand 9

IL2RB

Interleukin-2 Receptor Subunit Beta

AMPK

Adenosine monophosphate-activated protein kinase

SREBP1c

Sterol regulatory element-binding protein 1c

Authors’ contributions

H.MA (Mona A Hegazy) conceptualized and study design, A.D (Dalia Abd el Fattah) did all the epidemiological parts.M. SS (Samar Saad Mohamed), F.M (Mona Fathy), H. S (Nora Hassan) performed the laboratory analysis. All the authors H.MA(Mona A Hegazy), M.SS (Samar Saad Mohamed), S.EH (Eman H Saad), A.A ( [Ahmed Abdelghani](https:/pubmed.ncbi.nlm.nih.gov/?term=Abdelghani+A&cauthor_id=38872989) ), A.D (Dalia Abd el Fattah), M.MAE (Mohamed Ahmed Elsayed Mekki), FM (Mona Fathy), H.N (Nora Hassan) & A.O ( [Omar Ashoush](https:/pubmed.ncbi.nlm.nih.gov/?term=Ashoush+O&cauthor_id=38872989) )  share equally in data collection and writing the manuscript.

Funding

Non.

Data availability

Data will be available on reasonable request to the corresponding author. Individual participant data will be shared in data sets in a de-identified and anonymized format.

Declarations

Ethics approval and consent to participate

Ethics statement: Faculty of Medicine Research Ethics Committee at Cairo University approved the study with approval number (MS-298-2023).

Consent for publication

Written informed consent was taken from all contributing authors.

Competing interests

The authors declare no competing interests.