Comparison of the diagnostic performance of alanine aminotransferase and ultrasonography for discrimination of liver biopsy proven metabolic associated steatotic liver disease: a ROC curve analysis

Abstract

Background

MASLD is an increasingly prevalent cause of chronic liver disease worldwide, requiring accurate and accessible non-invasive diagnostic tools as its gold standard is liver biopsy. This study aimed to assess the diagnostic performance of ultrasound and alanine transaminase using liver biopsy as the reference standard to identify MASLD.

Method

Facility based cross-sectional study design was used. Sociodemographic data using a structured questionnaire, fatty liver diagnosis using imaging ultrasonography and ALT test using fasting venous blood were assessed from the study participants. The characteristics of study participants were assessed using chi-square and independent t-test. Ultrasound and ALT results were evaluated by ROC curve analysis and the AUC values compared using the DeLong test. Their sensitivity and specificity were analysed and cut-off value was determined using Youden index formula. The AUROC values were checked for age and sex with p-value < 0.05 taken for statistical significance.

Results

The AUC values of ALT and US were 0.91 (95% CI 0.87–0.95) and 0.83 (95% CI 0.76–0.9), respectively with (p ≤ 0.001). The sensitivity and specificity of US were 82.02% and 96.6% at moderate and severe MASLD status. The sensitivity and specificity for ALT were 68.54% and 97.06% at the cut-off value of 34 IU/L.

Conclusion

Ultrasound was the prioritized technique for the detection for MASLD, while ALT also had a good discriminating ability of MASLD. Both show highest discriminating ability of MASLD in male and female.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12902-026-02352-5.

Introduction

Metabolic dysfunction-associated steatotic liver disease (MASLD), previously termed non-alcoholic fatty liver disease (NAFLD), is defined as steatotic liver disease in the presence of one or more cardio metabolic risk factor(s) and the absence of excessive harmful alcohol intake [1]. MASLD is now the most common form of liver disease in developed countries with the global prevalence of 20%-30% and in patients with type 2 diabetes mellitus (T2DM), the prevalence is even between 55 and 80% [2]. Different studies indicated that the burden of MASLD among T2DM patients increased in Ethiopia [3–5].

MASLD has become a leading cause of morbidity and mortality among patients with T2DM due to hepatic complications like cirrhosis and hepatocellular carcinoma (HCC) [5]. MASLD also associated with extra hepatic conditions including high risk of developing cardiovascular disease [6]. The pattern of dyslipidemia was found higher in MASLD [7] and the total and regional fat-to-muscle ratio was found a good predictor of MASLD including sex and region specific patterns [8].

Liver biopsy is the gold standard for the identification and histological assessment of MASLD. The risks of liver biopsy, however, are not negligible. The invasiveness makes liver biopsy less ideal for all patients, especially given the high prevalence of MASLD. Because of these limitations, several noninvasive imaging and biomarker tests have been assessed [9]. Abdominal ultrasound (US) is used for screening asymptomatic patients with an incidental elevation of liver enzymes. It is the most common method for the qualitative assessment of hepatic steatosis because of its ready availability and affordability [10]. On US, a steatosis features include higher echogenicity than renal cortex and splenic parenchyma owing to intracellular fat vacuole accumulation, hepatomegaly, and intra-hepatic vascular blurring [11]. However, US cannot detect small amounts of hepatic steatosis when more than 20% of hepatocytes contain histologically visible fat droplets, with a reported moderate sensitivity and specificity [12].

Serum biomarker alanine aminotransferase (ALT) have been shown to be elevated in patients with MASLD/MASH, although normal aminotransferase levels do not rule out the possibility of steatosis or MASH; patients with advanced MASLD have been reported to have normal ALT levels [13]. The EASL–EASD–EASO NAFLD guidelines recommend screening in all patients with the metabolic syndrome using liver function tests (LFTs) and this guidelines recommend that all patients with elevated LFTs should be referred to the hepatologist [14].

MASLD often progresses without symptoms in its early stages, by the time symptoms appear patients may already have advanced fibrosis or cirrhosis. Early detection is crucial in preventing MASLD from progressing to advanced stages, where treatment becomes more complex [15]. There is currently no approved therapy for MASLD, although several drugs are in advanced stages of development. Because of the complex pathophysiology combination treatment is likely to be required for many patients including healthy lifestyle and weight reduction [16]. The notable risks of liver biopsy, sampling error, and inter- and intraobserver variability make liver biopsy less than ideal for all patients. This pave the way for the emerging of numerous imaging and blood-based tests, while all of them have dynamic diagnostic performance, sensitivity and specificity. Therefore, the study aimed to evaluate the diagnostic performance of ALT and US for the diagnosis of MASLD and identify its associated factors, which may provide evidence for non-invasive reliable diagnostic alternatives.

Methods and materials

Study design, period, area, and setting

This study was conducted at Debre Tabor Comprehensive Specialized Hospital, located in Debre Tabor town, found in the Amhara region, which is found 667 km from the capital city, Addis Ababa. A facility based cross-sectional study was conducted to determine the diagnostic performance of ALT and US for the diagnosis of MASLD among T2DM patients at Debre Tabor Comprehensive Specialized Hospital, North Central Ethiopia, from April 03 to August 27, 2025.

Study population

The study population included T2DM patients who underwent liver biopsy for evaluation of fatty liver, who visited Debre Tabor Comprehensive Specialized Hospital during the study period, and who fulfilled the eligibility criteria.

Eligibility criteria

Inclusion criteria

All adult (≥ 18 years) T2DM patients who underwent liver biopsy for evaluation of fatty liver, who never had liver malignancy or transplantation, who were willing to participate in this study without an excess alcohol drinking habit.

Exclusion criteria

People with known hepatic disease or inflammation, people who took drug that can cause liver toxicity, who had evidence of liver injury at the time of assessment and pregnant women were excluded from this study.

Sample size determination and sampling technique

Initially, the sample size was not established; instead, all T2DM patients who had a liver biopsy for the assessment of fatty liver were screened for eligibility, and 123 patients who met those requirements were taken to the US and blood testing.

Study variables

Outcome variable

• Liver biopsy.

Dependent variables

• Ultrasound result.

• Alanine Transaminase test result.

Data collection process and laboratory methods

Adult (≥ 18 years) T2DM Patients who underwent liver biopsy were recruited after the purpose of the study had been explained to them and they provided written informed consent. Structured questionnaire was used collect the socio-demographic data. An experienced medical laboratory technologist collected 5mL venous blood from T2DM patients, and the serum was used for testing of FBS, Liver enzymes, and lipid profile using Siemens Dimension clinical chemistry analyzer.

Liver biopsy

Liver biopsy for MASLD was performed in a hospital or radiology unit under ultrasound guidance to precisely localize the liver and avoid nearby organs and to improve accuracy and safety. After local anesthesia a needle is inserted through the right upper abdomen into the liver in a quick in-and-out motion lasting only seconds. An adequate specimen 1.5–2 cm long that contains at least 10–11 portal tracts was taken. The biopsy is then graded and NAS score was calculated. Grading reflects disease activity and includes steatosis (grade 0: <5%, grade 1: 5–33%, grade 2: 34–66%, grade 3: >66%).

Ultrasound

A radiologist conducted an ultrasonography examination. The basic sign for steatosis is the increased echogenicity of the liver parenchyma in comparison to the cortex of the right kidney because intracellular accumulation of fat vacuoles reflects the ultrasound beam. The classification of steatosis is usually graded as follows: grade 0: normal echogenicity of the right liver lobe in comparison with the cortex of the right kidney; grade 1: slight, diffuse increase in fine echoes in liver parenchyma with normal visualization of diaphragm and intrahepatic vessel borders; grade 2: moderate, diffuse increase in fine echoes with slightly impaired visualization of intrahepatic vessels and diaphragm; grade 3: marked increase in fine echoes with poor or non-visualization of the intrahepatic vessel borders, diaphragm, and posterior right lobe of the liver.

Data quality assurance

The questionnaire was first prepared for this study in English and then translated into Amharic and then translated back into English to check its consistency. (See Supplementary Material 1) The data were checked for completeness. The visual inspection of probes and cables, brightness contrast, system start up self-test, check for artifacts were performed for the US. Proper sample labelling, adequacy, proper container, haemolysis, and proper volume were checked for blood markers measurement. The pre-analytical, analytical, and post-analytical stage of laboratory analysis were performed following quality assurance techniques.

Data analysis and interpretation

Data were entered using EpiData v. 3.1, and exported into Stata v.17 for further editing and analysis. The Chi-square test and independent t-test were used to compare differences in the general characteristics of the study participants according to the presence of MASLD using the gold standard liver biopsy. A receiver operating characteristic (ROC) curve was drawn and the optimal cut-off point was determined. Sensitivity and specificity were calculated. The area under the ROC curve (AUROC) was used to evaluate the diagnostic accuracy. Statistical significance was set at a p-value < 0.05.

Results

Characteristics of study participants

In this study, 123 study participants were included and the prevalence of MASLD using liver biopsy 89 (72.36%). The prevalence was also determined using US and ALT and become 61(49.59%) and 73(59.35%) respectively (Fig. 1). The mean ± SD of age and DM duration of the participants were higher in MASLD positive patients. Majority of females 55(84.61%) and people with no exercise habit 68(79.06%) had MASLD (Table 1).

Fig. 1.

Prevalence of MASLD using different diagnostic mechanisms

Table 1.

Characteristics of study participants who underwent liver biopsy (n = 123)

Variables (n = 123)	MASLD by liver biopsy		p-value
	No (n = 34)	Yes (n = 89)
Age	47.55 ± 15.47	48.1 ± 13.4	0.047
Sex
Male	24	34	0.001
Female	10	55
Exercise habit
Yes	16	21	0.011
No	18	68
DM duration (years)	3.64 ± 2.55	6.5 ± 5.5	0.037
FBS (70–100 mg/dl)	169.05 ± 46.04	244 ± 109.17	0.001
TG (< 150 mg/dl)	140.9 ± 65.49	178.5 ± 99.46	0.04
TC (< 200 mg/dl)	136.17 ± 40.46	163.2 ± 51.47	0.0067
DB (0–3 mg/dl)	0.11 ± 0.09	0.07 ± 0.036	0.0004
ALT (0–56 IU/L)	25.1 ± 5.18	42.7 ± 21.5	0.000

ROC analysis of US and ALT against liver biopsy

In the AUROC analysis (Fig. 2) both US and ALT showed an acceptable performance index of > 0.7. Of these, US showed the best performance, with an AUROC of 0.91 (95% CI: 0.87–0.95). De Long nonparametric test was performed to compare the AUROC which showed statistically significant (p < 0.05). The sensitivity and specificity were 82.02% and 96.99% respectively for moderate to severe MASLD stages. The optimal cut-off value of ALT was calculated using Youden index formula and the AUROC of ALT at 34 IU/L was 0.83 (95% CI: 0.76–0.901), and the sensitivity and specificity were 68.54%, 97.06% respectively (Table 2).

Fig. 2.

The ROC curve and AUC of US and ALT against liver biopsy for MASLD diagnosis

Table 2.

Diagnostic performance of US and ALT for detecting MASLD (n = 123)

Variables	AUROC	p-value	Sensitivity	Specificity	PPV	NPV	Prevalence
US	0.91(0.87–0.95)	< 0.0001	82.02%	96.99%	82(72.5–89.4)	96.01(89.7–100)	59.3(50.1–68.1)
ALT	0.83(0.76–0.901)	< 0.0001	68.54%	91.06%	67.4(56.7–77)	97.1(84.7–99.9)	49.6(40.5–58.8)

Diagnostic accuracy for MASLD by sex

Table 3 shows the results of the subgroup analysis according to sex. The AUROC of US was 0.95 (95% CI: 0.9–1) in men and 0.88 (95% CI: 0.82–0.93) in women. Men showed better results than women. On the other hand, ALT showed more accurate results for women than men (Table 3).

Table 3.

Sex differences in diagnostic accuracy of ALT and US (n = 123)

	AUROC (95% CI) for male	p-value	AUROC (95% CI) for female	p-value
US	0.95 (0.9-1)	< 0.001	0.88 (0.82–0.93)	< 0.001
ALT	0.81 (0.69–0.92)	< 0.001	0.85 (0.74–0.96)	< 0.001

Diagnostic accuracy for MASLD by age

The performance of each index was assessed according to age group classified by the mean age of the study participants. The AUROC values of ALT were 0.81 for those aged below the mean age (47.9). US showed consistent AUROC values, for both age groups (Table 4).

Table 4.

Diagnostic accuracy of ALT and US in different age groups (n = 123)

	AUROC (95% CI) for age group < mean (47.95)	p-value	AUROC (95% CI) for age group ≥mean (47.95)	p-value
US	0.9 (0.85-0.96)	<0.001	0.91 (0.85-0.97)	<0.001
ALT	0.81 (0.72–0.91)	< 0.001	0.85 (0.75–0.95)	< 0.001

Discussion

There has been a rise in the prevalence of MASLD, paralleling a worldwide increase in DM and metabolic syndrome. However, it has several drawbacks, a liver biopsy remains the gold standard for making a definitive diagnosis [17]. In the future, improved understanding of the pathogenesis of MASH and new technologies may contribute to the diagnostic process and provide reliable, non-invasive alternatives to liver biopsy. Blood chemistry shows mild elevation of transaminases, and also other evidence for liver dysfunction in the cirrhotic stage [18] .In our study the prevalence of MASLD was different using liver biopsy 89(72.36), US 61(49.59), and ALT 73(59.35). This is consistent with the evidence that liver biopsy remains the gold standard and can detect mild steatosis missed by non-invasive techniques showed in different researches [19, 20]. The lower prevalence in US and ALT reflects the limitation in diagnostic performance particularly for mild steatosis.

The study found US and ALT had acceptable ability to differentiate MASLD with the AUC values of 0.91 and 0.83 respectively. Another study showed AUC values of US and ALT (0.70 and 0.74) which has fair ability to differentiate MASLD [21]. This difference may be caused due to the difference in the study population, which our study focused on adult T2DM patients and the other study focus on pediatrics, variations in disease severity and progression, metabolic burden, and methodological factors while utilizing both methods for MASLD detection. The other determinant may be the methods used for the reference value. This study used liver biopsy as a reference but the other used another imaging technique. A study showed predictive performance of ALT combined with AST and γ-glutamyl transferase and found AUROC of 0.822 (CI: 0.869–0.891) [22]. Another study conducted in Turkey [23] showed a consistent AUC value for ALT 0.817 (CI: 0.721–0.913) and this similarity may be explained by shared MASLD pathophysiology where hepatocellular injury leads to ALT release and global metabolic risk trends which may contribute to similar patterns of liver injury and ALT elevation. In addition the similarity of the study group also shared roles.

An excellent ability of differentiating MASLD was showed by a study which has a pooled AUC values of 0.93 which had similar values with the current study [24]. This similarity may be attributed to the consistent diagnostic performance of US across different population and study settings. Consistency with the pooled prevalence strengthens the reliability of US as non-invasive tool for MASLD detection. In our study, the combined model incorporating ALT and US findings did not demonstrate improved diagnostic performance. This may be partly attributed to the limited sample size of the study that can leads to poor model performance [25, 26]. The AUROC for the combined model was lower than that of ALT and US alone, suggesting no additive discriminatory value from integrating both variables.

Thus, to increase diagnostic accuracy, it is necessary to use an algorithm that considers the characteristics of the population. Considering the above results, applying customized models for age and sex is necessary. It seems reasonable to consider a specific cut-off value rather than uniformly applying the established cut-off value. Our study showed a good diagnostic accuracy in both male and female. This finding is in line with another study conducted in Chile [27] and Japan [28]. The similarity may be explained by the fact that the underlying hepatic fat accumulation and structural liver changes detected by US in both male and female, allowing comparable diagnostic accuracy. Also the result indicates the diagnostic performance of ALT is relatively stable between males and females, supporting their performance as effective tools for MASLD detection in diverse population.

Conclusion and recommendation

In summary, we investigated the diagnostic performance of US and ALT for identifying MASLD using liver biopsy. The area under the receiver operating characteristic (AUROC) curve was used to validate the diagnostic accuracy of MASLD. Both showed good performance for detecting MASLD in older patients. Of the two, US was the best at predicting MASLD, with the highest AUC (0.91). The sensitivity and specificity of ALT for diagnosing MASLD were 68.54% and 97.06%. The optimal cut-off value was 34. Diagnosing MASLD using the liver biopsy is an invasive technique which is difficult for applying routinely. Alternating non-invasive techniques is crucial for MASLD as its burden is increased with increasing of predisposing factors such as T2DM and other metabolic syndromes. Paying adequate attention by the stakeholders towards the non-communicable era and efforts should be made to standardize the non-invasive techniques to improve consistency, reproducibility, and comparability of results. It is recommended to conduct further researches in a large scale and diverse population to figure out the diagnostic abilities of liver biomarkers and US for MASLD and to increase generalizability of the findings.

Strength and limitation of the study

The strength of this study is, it performed ROC analysis which allowed evaluation of diagnostic performance and determination of optimal cut-offs using widely available diagnostic tools that can increase its practical applicability using the gold standard method as a reference. Although, it included the feasible tools, it is widely subjected for operator skill, equipment quality, and technical factors and its small sample size may affect its generalizability to the general population.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ALT

Alanine transaminase

AUC

Area under the curve

AUROC

Area under receiver operating characteristics curve

LFT

Liver function test

MASLD

Metabolic Associated steatotic Liver Disease

NAFLD

Non-alcoholic fatty liver disease

ROC

Receiver operating characteristics

T2DM

Type 2 Diabetes Mellitus

US

Ultrasound

Author contributions

MF : conceived, designed, performed the laboratory tests and statistical analysis, interpreted the data, and wrote the draft manuscript. EE, GM, RA, BK, AB, TK, and BE : collected the data, performed the laboratory tests and statistical analysis, and wrote the draft manuscript. All authors read and approved the final version of the manuscript and the corresponding author (Mahider Shimelis Feyisa) have full access of the data and takes complete responsibility for the integrity of the data and accuracy of the data analysis.

Funding

The authors have not received a specific grant for this research from any funding agency in the public, commercial or non-profit sectors.

Data availability

Upon the corresponding author’s request, all the necessary information is made available.

Declarations

Ethical approval and consent

The study was conducted in accordance with Declaration of Helsinki and after obtaining ethical clearance from the internal ethical review committee of Debre Tabor University, College of Health Science, Medical Laboratory Sciences Education and Service Directorate (IRB) (Reference Number: 179/2024). Participants who were willing to participate in the study provided a written informed consent and only when a participant was willing and ready to take part in the research did the data collection process start. Furthermore, relevant guidelines and regulations were followed in all process.

Consent for publication

Not applicable.

Competing interests

The authors report no conflicts of interest for this work.