Assessment of metabolic biomarkers associated with hepatic steatosis in Crohn’s disease

Abstract

BACKGROUND

Steatotic liver disease (SLD) is the most frequent hepatic manifestation in inflammatory bowel disease. The contribution of intestinal inflammatory activity and hepatokines to steatosis development in Crohn’s disease (CD) remains unclear. We aimed to evaluate the prevalence of steatosis according to CD activity, analyze metabolic and hepatokine profiles, and assess the performance of non-invasive imaging techniques.

AIM

To investigate SLD, associated biomarkers, and diagnostic imaging accuracy in CD.

METHODS

Seventy-two subjects were included: Active CD (n = 19), CD in sustained deep remission (n = 24), and healthy controls (n = 29). Patients with diabetes mellitus, chronic liver disease, or significant alcohol consumption were excluded. CD activity was assessed using, among other parameters, intestinal ultrasound. SLD was assessed by ultrasound and controlled attenuation parameter (CAP). Liver stiffness was measured by transient elastography and two-dimensional shear wave elastography. Serum angiopoietin-like protein 3 (ANGPTL3), ANGPTL6, fibroblast growth factor 19 (FGF19), and FGF21 levels were determined. Metabolic assessment included lipid profile, homeostasis model assessment of insulin resistance, and glycated hemoglobin (HbA1c).

RESULTS

Overall prevalence of SLD was 33% (24/72): 53% in active CD, 30% in CD remission, and 21% in controls (P = 0.04). Among patients with SLD, HbA1c levels were significantly higher in active CD (31.4 mmol/mol) compared to CD in remission (17.8 mmol/mol) and controls (5.8 mmol/mol) (P = 0.007). No significant differences in liver enzymes were observed among groups with steatosis. Patients with steatosis showed higher levels of ANGPTL6 (124.5 ng/mL vs 112.8 ng/mL; P = 0.04) and FGF21 (0.12 ng/mL vs 0.06 ng/mL; P = 0.01) compared to subjects without steatosis, although no differences were observed among the different cohorts. No significant differences were found in ANGPTL3, ANGPTL4, or FGF19, either according to the presence of steatosis or between cohorts. CAP values were significantly higher in ultrasound-detected SLD (P = 0.0001), with area under the curve (AUC) ranging. CAP values were higher in patients with active CD compared to remission CD and controls; however, CAP demonstrated adequate diagnostic accuracy for steatosis across all three groups, with AUC values (0.76-0.82).

CONCLUSION

Inflammatory activity in CD is associated with a higher prevalence of SLD and subclinical metabolic alterations. ANGPTL6 and FGF21 may serve as potential biomarkers of SLD in CD. CAP appears to be a reliable non-invasive tool, although inflammatory status should be considered.

Key Words: Crohn’s disease; Metabolic dysfunction-associated steatotic liver disease; Hepatic steatosis; Hepatokines; Angiopoietin-like protein 6; Fibroblast growth factor 21; Controlled attenuation parameter; Intestinal inflammation

Core Tip: Active Crohn’s disease is associated with a higher prevalence of steatosis and increased angiopoietin-like protein 6 and fibroblast growth factor 21 levels, suggesting inflammation-related metabolic alterations. Controlled attenuation parameter may serve as a reliable non-invasive tool for steatosis detection in Crohn’s disease, although inflammatory activity should be considered in its interpretation.

INTRODUCTION

Metabolic-associated fatty liver disease, currently referred to as metabolic dysfunction-associated steatotic liver disease (MASLD), represents one of the leading causes of chronic liver disease and has become a major determinant in the development of cirrhosis in Western countries. In the context of inflammatory bowel disease (IBD), steatotic liver disease (SLD) is also the most frequently detected hepatic alteration, with an incidence of approximately 40% reported in some series[1,2].

It is important to differentiate MASLD from the broader entity of SLD. While MASLD is defined by the presence of metabolic dysfunction, SLD includes hepatic steatosis arising from other mechanisms. In IBD, steatosis may be driven predominantly by chronic systemic inflammation rather than classical metabolic risk factors, making SLD a more appropriate term in this context.

Although the molecular mechanisms involved in the onset and progression of MASLD and SLD are not yet fully understood, there is broad consensus that the initial events are closely related to obesity and insulin resistance (IR), both at the hepatic and adipose tissue levels. In addition, processes such as oxidative stress, mitochondrial dysfunction, lipotoxicity, and gut-derived endotoxinemia have been linked to disease progression toward more advanced stages[3,4].

Previous studies have identified intestinal inflammatory activity, corticosteroid use, biologic therapies, and intestinal resection as potential risk factors associated with SLD in patients with IBD. However, the precise mechanisms by which intestinal inflammation promotes hepatic fat accumulation remain unclear. Proposed hypotheses include the involvement of pro-inflammatory cytokines, alterations in lipoprotein lipase function, and peripheral IR as potential mediators of this process[2,4].

In this context, specific hepatokines and growth factors have gained increasing attention as potential mediators linking IBD and SLD. Among these, angiopoietin-like proteins (ANGPTLs) including ANGPTL3, ANGPTL4, and ANGPTL6 as well as fibroblast growth factors (FGFs), particularly FGF19 and FGF21, have emerged as key regulators of lipid metabolism, bile acid homeostasis, and insulin sensitivity[5-11]. Dysregulation of ANGPTL3, ANGPTL4, and ANGPTL6 has been associated with abnormalities in triglyceride catabolism and lipid distribution, thereby favoring a pro-steatogenic environment.

ANGPTL3, in particularly, is a hepatokine synthesized almost exclusively in the liver that inhibits lipoprotein lipase and endothelial lipase, modulating the clearance of triglyceride-rich lipoproteins and systemic lipid metabolism[6,7]. Given its central role in lipid metabolism, ANGPTL3 has been proposed as a therapeutic target for dyslipidemia and potentially for MASLD[8-10]. However, a direct role of ANGPTL3 in hepatic steatosis among patients with IBD has not yet been demonstrated, although genetic and proteomic studies have identified it as a potential target and biomarker in IBD[12-14].

Similarly, FGFs such as FGF19 and FGF21 act as metabolic and inflammatory hormones, modulating bile acid synthesis, fatty acid oxidation, and responses to metabolic stress[15-17]. In patients with IBD, chronic inflammation and alterations of the gut-liver axis may significantly modify the levels and activity of these hepatokines, thereby contributing to the development and progression of SLD and representing a potential pathophysiological link as well as an emerging therapeutic target.

With regard to the FGF19-mediated gut-liver axis, studies in patients with Crohn’s disease (CD) have shown that higher circulating levels of FGF19 particularly in those treated with biologic therapies such as anti-tumor necrosis factor agents are associated with a lower degree of hepatic steatosis, suggesting a potential hepatoprotective effect. Conversely, reduced FGF19 levels in conditions such as bile acid malabsorption related diarrhea have been associated with an increased risk of SLD, further underscoring the importance of FGF19 in hepatic metabolic regulation. Accordingly, recent reviews have highlighted this enterohepatic circuit as one of the main pathophysiological mechanisms underlying the coexistence of IBD and SLD[16,18].

In parallel, FGF21 has emerged as a hormone with a prominent role in energy metabolism and stress responses. In IBD, circulating FGF21 levels have been primarily associated with the degree of inflammation and nutritional status; however, studies simultaneously evaluating this hepatokine and the presence of SLD in IBD cohorts are lacking[19-21]. Nevertheless, in the field of metabolic liver disease, the evidence is robust: FGF21 and its therapeutic analogues exerts beneficial effects on hepatic lipid accumulation, inflammation, and fibrosis.

Although the prevalence of SLD in IBD has been increasingly recognized, the biomarker profile specifically associated with SLD in CD remains poorly characterized. In particular, it is still unclear whether hepatokines described in MASLD in the general population behave similarly in the setting of CD, where chronic intestinal inflammation, altered gut-liver signaling, and previous exposure to multiple therapies may contribute to a distinct pathophysiological background.

Collectively, these data point to hepatokines as potential links between intestinal inflammation and metabolic liver injury, highlighting their value as candidate non-invasive biomarkers of SLD. Therefore, this study aimed to assess the diagnostic utility of hepatokines in patients with CD and to explore their contribution to the pathophysiology of SLD associated with IBD.

MATERIALS AND METHODS

Study population

Three cohorts of patients were prospectively included: (1) Patients with CD, involving the small bowel and/or colon, with evidence of clinical, biochemical, and imaging (ultrasound and/or endoscopic) inflammatory activity prior to the initiation of biologic therapy (cohort 1); (2) Patients with CD in sustained deep remission, defined by combined biochemical, clinical, and radiologic remission for at least one year (cohort 2); and (3) A control group of healthy individuals, provided they met none of the study exclusion criteria (cohort 3). Patients were recruited between April 2024 and December 2024, with subsequent follow-up of those requiring it according to the study protocol.

Patients and healthy controls fulfilling the inclusion criteria were age- and sex-matched across groups. Healthy controls were recruited from hospital staff and community volunteers during the same study period. To minimize potential selection bias related to metabolic syndrome and to avoid overrepresentation of individuals with a high baseline risk of SLD, no more than 30% of the study population was allowed to present two or more established metabolic risk factors (such as obesity or dyslipidemia). Given the high prevalence of SLD among patients with diabetes mellitus, individuals with a prior diagnosis of diabetes were excluded from all cohorts.

No formal a priori sample size calculation was performed because this was an exploratory, hypothesis-generating study; however, approximately 20-30 patients per group were included to help approximate normality assumptions. Missing data were handled using complete-case analysis for each specific comparison, and no imputation procedures were applied.

Assessment of CD activity

All patients underwent a comprehensive assessment of CD activity at study inclusion. Clinical activity was evaluated using Harvey-Bradshaw index (HBI). Biochemical activity was assessed by measuring fecal calprotectin, C-reactive protein (CRP), hemoglobin, and serum albumin. Intestinal activity was further assessed using intestinal ultrasound in all CD patients. In those with evidence of clinical (defined as HBI > 5), biochemical (defined as calprotectin levels > 150 μg/g) or ultrasonographic activity, endoscopic evaluation by ileocolonoscopy was performed, and inflammatory activity was determined using the Simple Endoscopic Score for CD. Patients included in the sustained deep remission cohort were required to have no evidence of clinical or biochemical activity during the preceding year, defined as HBI < 5, fecal calprotectin < 150 μg/g and CRP < 5 mg/L.

Metabolic and hepatic assessment

At inclusion, all participants underwent a structured evaluation of metabolic risk factors and the presence of SLD. Metabolic risk factors were assessed individually rather than by a formal metabolic syndrome definition. The variables recorded for this purpose were body mass index, previously established diagnosis of hypertension and dyslipidemia. Obesity was considered when body mass index was ≥ 30 kg/m².

Serum levels of ANGPTL3, ANGPTL6, FGF19, and FGF21 were measured. A complete lipid profile including total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, and triglycerides was obtained, together with liver biochemistry [total and direct bilirubin, alanine aminotransferase (ALT), aspartate aminotransferase, gamma-glutamyl transferase (GGT), and alkaline phosphatase]. IR was estimated using the homeostasis model assessment-IR index, calculated from fasting glucose and insulin levels. Glycated hemoglobin (HbA1c) was also measured in all patients.

Imaging assessment of hepatic steatosis and fibrosis

High-resolution ultrasound examinations were performed using a commercially available system (Aplio i800, Canon Medical Systems, Otawara, Japan). All scans were conducted by operators with extensive experience in hepatic ultrasonography. Examinations followed a standardized liver exploration protocol in all participants to ensure consistency and reproducibility.

Hepatic steatosis was assessed using B-mode ultrasonography and defined as a dichotomous outcome (present/absent). Conventional abdominal ultrasonography was considered the reference (gold-standard) imaging modality for the diagnosis of hepatic steatosis in this study. The diagnosis of steatosis was established when at least two of the following criteria were present: (1) Increased echogenicity of the hepatic parenchyma relative to the renal cortex or spleen; and (2) Additional supportive features including posterior beam attenuation, blurring of intrahepatic vascular margins, poor visualization of the gallbladder wall, or impaired delineation of the diaphragm.

Additionally, transient elastography (FibroScan^®) was performed to measure liver stiffness and to quantify hepatic steatosis using the controlled attenuation parameter (CAP^®). Hepatic steatosis was defined as CAP ≥ 248 dB/m and further graded according to established thresholds as mild (S1: ≥ 248 dB/m), moderate (S2: ≥ 268 dB/m), or severe (S3: ≥ 280 dB/m).

Advanced liver fibrosis was assessed using validated cut-off values: Liver stiffness < 8 kPa was considered to exclude advanced fibrosis (≤ F2), whereas values ≥ 12 kPa were considered indicative of advanced fibrosis (≥ F3). Ultrasound and transient elastography assessments were performed by operators blinded to laboratory findings.

Clinical and treatment data

Baseline demographic and clinical data were collected, including date of birth, sex, smoking status, alcohol consumption, comorbidities, and cardiovascular risk factors (body mass index, hypertension, and dyslipidemia). Disease-related variables were also recorded, including prior treatments and surgical resections.

For patients initiating biologic therapy for control of intestinal inflammation, the specific agent (infliximab, adalimumab, ustekinumab, vedolizumab, or risankizumab), dosage, route of administration, and initiation date were documented. Use of Janus kinase inhibitors (tofacitinib or upadacitinib) and concomitant immunosuppressive therapy (azathioprine or mercaptopurine) was also recorded.

Patients who had received systemic corticosteroids within three months prior to screening or who were receiving methotrexate were excluded due to the potential steatogenic effects of these drugs. Patients initiating biologic therapy for indications other than luminal disease activity (e.g., extraintestinal immune-mediated conditions), those treated exclusively for perianal disease without luminal involvement, and those with ileal or ileocolonic resection within the previous year were also excluded.

Individuals with alcohol consumption exceeding 20 g/day in women or 30 g/day in men during the previous year were excluded, as were patients with concomitant liver diseases, including hepatitis B or C, primary biliary cholangitis, primary sclerosing cholangitis, Wilson’s disease, hemochromatosis, autoimmune hepatitis, alpha-1 antitrypsin deficiency, drug-induced liver injury, or biliary obstruction. Additionally, participants receiving or having received treatment with glucagon-like peptide-1 receptor agonists within the previous three months were excluded, due to their known effects on hepatic steatosis and metabolic parameters. Subjects who had experienced significant recent weight loss (≥ 5% of body weight within the preceding six months) were also excluded, as this may independently influence liver fat content. The study was approved by the Ethics Committee of Hospital Universitario La Paz (Madrid) in April 2024 (HULP code: No. PI-6155). Written informed consent was obtained from all participants before study inclusion.

Statistical analysis

A descriptive analysis of baseline characteristics and disease-related variables was performed. Continuous variables were expressed as mean ± SD, while categorical variables were reported as n (%) with 95% confidence intervals. Categorical variables were compared using the χ² test, and quantitative variables were analyzed using Student’s t-test when normally distributed; otherwise, appropriate non-parametric tests were applied. Normality was assessed using the Kolmogorov-Smirnov test. Receiver operating characteristic (ROC) analysis evaluated the diagnostic performance of CAP across groups, while logistic regression was used to explore the association between study group and ultrasound-detected steatosis after adjustment for age, sex, dyslipidemia, and hypertension. In addition, a linear regression model was fitted to evaluate the potential interaction between CD activity status and steatosis on circulating ANGPTL6 levels, adjusting for age and sex. Statistical significance was set at P < 0.05. All statistical analyses were performed using Stata version 16 (StataCorp, College Station, TX, United States).

RESULTS

A total of 72 subjects were included: 19 patients with active CD (27%, cohort 1), 24 patients with CD in sustained deep remission (33%, cohort 2), and 29 healthy controls (40%, cohort 3). The age range of the overall study population was 18-78 years. Although no statistically significant differences in age were observed between groups, sex distribution differed significantly across cohorts (P = 0.04), with a higher proportion of women in the healthy control group. Therefore, dyslipidemia was more prevalent among patients with CD in remission compared with those with active disease [29.1% (7/24) vs 5.2% (1/19); P = 0.002]. Baseline characteristics are summarized in Table 1. Absence of clinical, biochemical, and ultrasonographic activity was confirmed in the CD remission group. These data are summarized in Table 2.

Table 1 Baseline characteristics, n (%).

Variable	CD remission	Active CD	Healthy controls	Statistical significance
Sex				0.04
Female	12 (30)	7 (17.5)	21 (52.5)
Male	12 (37.5)	12 (37.5)	8 (25)
Smoking status				0.007
No	10 (25)	6 (15)	24 (60)
Yes	6 (37.5)	5 (31.25)	5 (31.5)
Former	7 (63.6)	4 (36.4)	0 (0)
Weight (kg), median IQR	76.3 (62-84)	77.25 (65-85)	76.5 (61-90)	0.3
Height (cm), median IQR	171.5 (163.5-179)	172 (167-175)	162.5 (158-174)	0.2
Dyslipidemia	7 (29.1)	1 (5.2)	0 (0)	0.002
Hypertension	3 (37.5)	3 (37.5)	2 (25)	0.6
Prior diagnosis of MASLD	1 (4.1)	1 (5.2)	0 (0)	0.4

CD: Crohn’s disease; MASLD: Metabolic dysfunction-associated steatotic liver disease; IQR: Interquartile range.

Table 2 Baseline inflammatory bowel disease activity and disease-related characteristics, n (%)/median interquartile range.

Variable	CD remission	Active CD	Healthy controls	Statistical significance
Previous IBD treatment	18/24 (75)	14/19 (73.7)
Clinical activity (HBI)	0 (0)	3 (0-6)		< 0.0001
Fecal calprotectin (μg/g)	45.1 (20.7-63)	331 (209-418)		< 0.0001
C-reactive protein (mg/L)	0.8 (0.5-1.6)	4.8 (1.4-8.4)	0.5 (0.5-1.9)	< 0.0001
Hemoglobin (g/dL)	14.3 (13.9-15.8)	14.6 (13.9-15.8)	14.5 (13.8-15.2)	0.07
Albumin (g/dL)	4.4 (4.2-4.5)	4.5 (4.5-4.6)	4.7 (4.4-4.8)	0.3
Endoscopic activity (SES-CD)	12 (5-13)	0 (0-3)
Ultrasound: Bowel wall thickness (mm)	2 (2-2)	4.8 (4.5-5.6)
Ultrasound: Hyperemia (Limberg score)	0 (0)	3 (2-3)
Loss of wall stratification	0 (0)	2 (13)		0.07
Mesenteric fat proliferation	0 (0)	4 (26)		0.01

CD: Crohn’s disease; IBD: Inflammatory bowel disease; HBI: Harvey-Bradshaw index; SES-CD: Simple Endoscopic Score for Crohn’s disease.

Overall, 33% of patients (n = 24) showed evidence of hepatic steatosis on ultrasound. The prevalence of steatosis by group was 30% (n = 7) in CD remission, 53% (n = 10) in active CD, and 21% (n = 6) in healthy controls (P = 0.04). A multivariate analysis was performed using the presence of ultrasound-detected steatosis (yes/no) as the dependent variable and to account for the imbalance in sex distribution and potential metabolic confounding, an adjusted logistic regression model including age, sex, dyslipidemia, body mass index, baseline weight and hypertension was performed. Baseline weight was the only variable significantly associated with steatosis (P = 0.006; odds ratio = 1.1). However, no significant differences in baseline weight were observed between groups. Sex was not independently associated with SLD in the adjusted model.

Serum biochemical parameters

Among patients with hepatic steatosis, no statistically significant differences were observed in lipid profiles, except for higher triglyceride levels in patients with CD, particularly those in sustained remission. No clinically or statistically relevant differences were observed in biochemical parameters among patients with SLD across the three cohorts (Table 3). Although no patients with diabetes mellitus were included, higher HbA1c levels (mmol/mol) were observed in patients with active CD and hepatic steatosis compared with those in remission and healthy controls (31.4 vs 17.8 vs 5.8; P = 0.007).

Table 3 Biochemical and metabolic parameters according to study groups (mean value).

Variable	CD remission	Active CD	Healthy controls	Statistical significance
Total cholesterol (mg/dL)	186	174	179	0.6
Triglycerides (mg/dL)	134	101	87	0.04
LDL cholesterol (mg/dL)	133	113	98	0.09
HDL cholesterol (mg/dL)	54	52	61	0.1
ALT (U/L)	30	21	24	0.06
AST (U/L)	22	20	17	0.08
GGT (U/L)	31	21	23	0.1
ALP (U/L)	80	87	89	0.01
Apolipoprotein B	101	85	78	0.1
HOMA index	2.54	1.45	2.98	0.09
HbA1c (mmol/mol)	17.7	35.5	5.8	0.02

CD: Crohn’s disease; LDL: Low-density lipoprotein; HDL: High-density lipoprotein; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; GGT: Gamma-glutamyl transferase; ALP: Alkaline phosphatase; HOMA: Homeostasis model assessment; HbA1c: Glycated hemoglobin.

The analytical profile of patients diagnosed with SLD was evaluated, and no significant differences were identified in liver biochemical parameters. However, higher GGT levels were observed in patients with CD in remission and hepatic steatosis compared with the other groups (76.6 vs 30.8 in active CD vs 24 in healthy controls; P = 0.4). Likewise, among patients with CD and SLD in both groups, higher mean ALT levels were identified compared with healthy controls (CD in remission: 41.1 vs active CD: 42.6 vs healthy controls: 23.4; P = 0.5). These results are summarized in Table 4.

Table 4 Liver biochemical profile in patients with ultrasound-diagnosed hepatic steatosis (mean value).

Variable	CD remission	Active CD	Healthy controls	P value (ANOVA)
AST (UI/L)	26.6	27.3	18.6	0.5
ALT (UI/L)	41.1	42.6	23.4	0.5
LDL (mg/dL)	125	112.2	99.4	0.3
GGT (UI/L)	76.6	30.8	24	0.4
ALP (UI/L)	83.5	91.8	85.8	0.7
Bilirubin (mg/dL)	0.8	0.5	0.6	0.2

CD: Crohn’s disease; AST: Aspartate aminotransferase; ALT: Alanine aminotransferase; GGT: Gamma-glutamyl transferase; ALP: Alkaline phosphatase; LDL: Low-density lipoprotein; ANOVA: Analysis of variance.

Hepatokines

Regarding the study of hepatokines, patients with hepatic steatosis detected by ultrasound presented higher mean levels of ANGPTL6 (124.5 ng/mL vs 112.8 ng/mL; P = 0.04) and FGF21 (0.12 ng/mL vs 0.06 ng/mL; P = 0.01) (Figure 1). No clinically or statistically significant differences were observed in ANGPTL3, ANGPTL4, or FGF19 levels between patients with or without hepatic steatosis on ultrasound across the three cohorts.

Figure 1 Serum hepatokine levels according to hepatic steatosis. Serum levels of angiopoietin-like protein 6 (ANGPTL6) and fibroblast growth factor 21 (FGF21) in subjects with and without ultrasound-diagnosed steatotic liver disease (SLD). Patients with SLD showed significantly higher levels of ANGPTL6 and FGF21 No significant differences were observed for other hepatokines. ANGPTL6: Angiopoietin-like protein 6; FGF21: Fibroblast growth factor 21.

We also analyzed the behavior of these molecules exclusively in patients with ultrasound-diagnosed steatosis. No differences were observed in ANGPTL3 (CD remission 19.3 vs CD activity 18.6 vs healthy controls 19.2; P = 0.9). Regarding ANGPTL4, higher levels were observed in patients with CD activity (166.8) compared with patients in CD remission (134.7) and healthy controls (135.3), although these differences did not reach statistical significance. Finally, ANGPTL6 levels were lower in patients with CD remission and steatosis: 106 vs 134.7 in the other groups, without reaching statistical significance (P = 0.44).

In an exploratory linear regression model adjusted for age and sex, SLD remained associated with higher ANGPTL6 levels, whereas no significant interaction was observed between active CD and SLD on ANGPTL6 concentrations (P = 0.61). Similar results were obtained when the analysis was restricted to CD patients only, with no evidence of interaction between disease activity and steatosis (P = 0.87).

No statistically significant differences were identified among patients with ultrasound-detected steatosis across the different cohorts for FGF19 (mean 0.23 in CD remission, 0.12 in CD activity, and 0.14 in healthy controls; P = 0.45) or FGF21 (mean 0.14 in CD remission, 0.12 in CD activity, and 0.09 in healthy controls; P = 0.37). Differences in the behavior of ANGPTLs and FGFs between patients with CD from both cohorts and healthy controls were also investigated, with no clinically or statistically significant differences identified.

Liver stiffness measurements by transient elastography

Regarding the presence of fibrosis, no significant fibrosis was identified in any of the included patients, with no significant differences between the median values of the different cohorts. Median FibroScan^® values were 4.95 kPa (CD in remission), 4.1 kPa (active CD), and 4.5 kPa (healthy controls), while median values assessed by two-dimensional shear wave elastography were 4.1 kPa (CD in remission), 4.1 kPa (active CD), and 4.5 kPa (healthy controls).

CAP values were significantly higher in patients with steatosis detected by ultrasound (296.7 ± 75.0 dB/m vs 222.6 ± 44.9 dB/m; P = 0.0001). When comparing CAP values exclusively among patients with ultrasound-confirmed SLD, values were numerically higher in patients with CD activity (319.4 ± 84.8) compared with those in CD remission (288.9 ± 64.0) and healthy controls (271.2 ± 77.9); however, these differences did not reach statistical significance (P = 0.52). The area under the curve (AUC) was 0.76 (CD remission), 0.82 (CD activity), and 0.77 (healthy controls), with optimal CAP cut-off values of 234 dB/m (sensitivity 85.7%, specificity 60%), 250 dB/m (sensitivity 87.5%, specificity 78%), and 241 dB/m (sensitivity 80%, specificity 77%), respectively.

Hepatic steatosis severity was subsequently categorized using CAP measurements. Overall, 73% of the included patients presented moderate-to-severe steatosis (≥ 268 dB/m), while 27% had mild steatosis (≥ 248 -267dB/m). Severe steatosis was present in 22.7% of patients in remission, 35.3% of those with active disease, and 6.9% of healthy controls (P = 0.03).

When CAP was dichotomized into mild vs moderate/severe steatosis, no statistically significant differences in ANGPTL3, ANGPTL6, FGF19, or FGF21 circulating levels were observed across any of the study groups. Significantly higher ANGPTL4 levels were detected in patients with moderate/severe steatosis within the active CD group (195.8 vs 123.5; P = 0.04), whereas no differences were found in the CD remission or control groups.

Patients with moderate-to-severe steatosis exhibited higher HbA1c levels compared with those with mild steatosis (22.7 vs 9.1; P = 0.04). In subgroup analyses, this difference remained significant only among patients with active CD (34.8 vs 18; P = 0.05).

DISCUSSION

In our cohort, the prevalence of SLD diagnosed by ultrasound was significantly higher in patients with active CD compared with those in sustained deep remission and healthy controls (53%, 30%, and 21%, respectively; P = 0.04). This finding, consistent with previous reports, suggests that active intestinal inflammation constitutes an independent determinant of hepatic steatosis, even in the absence of classical metabolic risk factors. Despite higher dyslipidemia prevalence in remission, steatosis was highest in active CD, supporting an inflammation-driven contribution. Available literature, including recent studies using non-invasive techniques such as transient elastography and CAP, has described an increased prevalence of SLD in IBD particularly in CD and a potential modulatory effect of intestinal inflammation on hepatic lipid accumulation[1,2].

In line with previous studies indicating that SLD associated with IBD may present with minimal or absent biochemical abnormalities, no clinically relevant differences were observed in liver biochemistry including transaminases, GGT, alkaline phosphatase, and bilirubin between patients with or without CD who had ultrasound-confirmed hepatic steatosis.

Importantly, our results extend current evidence by showing that intestinal inflammatory activity is associated not only with a higher prevalence of steatosis but also with subtle metabolic alterations, such as significantly higher HbA1c levels in patients with active CD (P = 0.007). Although none of the participants met diagnostic criteria for diabetes mellitus, this finding suggests the presence of subclinical glycemic dysregulation, likely related to systemic inflammation in active CD and increased hepatic metabolic susceptibility. Taken together, these observations underscore that metabolic dysregulation rather than overt biochemical liver injury may represent the primary link between active intestinal inflammation and the development of SLD in IBD.

Regarding metabolic and hepatic biomarkers, our study identified higher circulating levels of ANGPTL6 and FGF21 in patients with hepatic steatosis, whereas ANGPTL3, ANGPTL4, and FGF19 did not show significant differences according to the presence of hepatic fat infiltration or across CD cohorts and healthy controls. These findings are consistent with growing evidence identifying ANGPTL6 as a hepatokine closely associated with metabolic dysfunction, IR, and progression of metabolic liver disease[11]. The increase in ANGPTL6 observed in our steatosis cohort may reflect a state of hepatic metabolic dysfunction driven by both intestinal inflammation and systemic lipid metabolism alterations.

Conversely, the lack of significant differences in ANGPTL3 and ANGPTL4 tempers their role as metabolic biomarkers of SLD in the specific context of IBD. Although both molecules are involved in triglyceride catabolism and have been proposed as therapeutic targets in dyslipidemia and non-alcoholic fatty liver disease (NAFLD)[7,8,22], our data do not support a direct association between their serum concentrations and SLD in CD. This finding aligns with recent multidisciplinary studies that, while identifying ANGPTL3 as a candidate molecule in certain IBD subgroups, have not confirmed its role in hepatic pathogenesis[14,16].

With respect to FGFs, no significant differences in FGF19 or FGF21 levels were observed across the three cohorts, although a trend toward lower FGF19 levels was noted in patients with active CD. This pattern is consistent with existing literature describing impairment of the FGF19 bile acid enterohepatic axis in inflammatory CD, particularly in the presence of diarrhea or bile acid malabsorption[16,18]. In contrast, the elevation of FGF21 in patients with steatosis supports its role as a hormone associated with hepatic metabolic stress, as widely reported in NAFLD/MASLD[23-25]. Our findings suggest that FGF21 may serve as a sensitive metabolic marker in IBD, although not specific to steatosis in the setting of CD.

One of the most relevant findings of the present study is the evaluation of non-invasive methods for hepatic steatosis assessment, such as CAP, in a specific clinical context such as IBD and particularly CD where intestinal inflammatory activity may influence the diagnostic performance of these techniques. Moreover, CAP-based stratification indicated that active CD was associated not only with a higher prevalence of steatosis but also with a greater proportion of moderate-to-severe cases compared with remission and control groups.

To date, CAP has not been specifically evaluated in CD cohorts, despite its increasing use in the non-invasive stratification of SLD. In our study, CAP demonstrated consistent diagnostic performance across all groups, with ROC AUC values ranging from 0.76 to 0.82, even in patients with active CD. Moreover, the variability in optimal cut-off values across clinical groups reinforces the hypothesis that intestinal inflammatory activity differentially affects quantitative attenuation-based techniques, highlighting the need for contextualized interpretation of non-invasive methods for hepatic steatosis assessment in IBD.

However, several aspects should be considered when interpreting these results. This was an exploratory cross-sectional study with prospectively collected data, and subgroup analyses were relatively small and potentially underpowered, particularly for detecting modest differences in hepatokines or liver biochemical parameters. Therefore, the absence of significant differences for several biomarkers should not be interpreted as definitive evidence of no association. In addition, five hepatokines were analyzed, and the nominally significant findings for ANGPTL6 and FGF21 should be interpreted with caution, as they were attenuated after correction for multiple testing, reinforcing the exploratory nature of these observations.

Overall, our findings do not support a specific mechanistic role for hepatokines in the development of SLD in CD. Nevertheless, alterations in markers such as ANGPTL6 and FGF21 may still reflect underlying metabolic disturbances and could have potential value as non-invasive biomarkers for the identification of steatosis. Further studies with larger sample sizes are warranted to confirm these findings and to better define their clinical and pathophysiological relevance in IBD.

Finally, the contribution of CAP as diagnostic tool in CD opens an important avenue for future research, particularly in a clinical setting where non-invasive and reproducible assessment of SLD is needed. The influence of intestinal inflammatory status on their diagnostic performance underscores the need for interpretation algorithms specifically adapted to patients with IBD.

CONCLUSION

Our results suggest that inflammatory activity in CD is associated with a higher burden of SLD and subtle metabolic alterations, including higher HbA1c levels even in the absence of diabetes mellitus. Although most hepatokines did not differ significantly between groups, ANGPTL6 and FGF21 emerged as potentially relevant exploratory biomarkers of SLD in this setting. Finally, CAP appears to be a useful non-invasive method for steatosis assessment in CD, although inflammatory status should be considered when interpreting its results.