We retrospectively identified inpatient ACLF patients who were admitted to the First Affiliated Hospital of Chongqing Medical University between October 2012 and July 2021. ACLF was diagnosed according to the criteria by Asia-Pacific Association for the Study of the Liver (APASL), which included jaundice [total bilirubin (TB) ≥ 5 mg/dL] and coagulation dysfunction [international normalized ratio (INR) ≥ 1.5], with the onset of ascites and/or hepatic encephalopathy (HE) within 4 weeks (1). Enrolled subjects were over 18 years of age, received abdominal CT examination within 2 weeks after admission, and were followed for a least 360 days. Exclusion criteria were (1) history of cirrhosis; (2) diagnosis of hepatocellular carcinoma (HCC) or other malignancies; (3) suffering from other disorders causing malnutrition such as tuberculosis, hyperthyroidism, or neuromuscular disease; (4) receiving long-term corticosteroid therapy; (5) death or undergoing liver transplant surgery.

This study was approved by the Ethics Committee of the First Affiliated Hospital of Chongqing Medical University (2022-K434). The requirement to obtain informed consent from patients was waived because the study was retrospective in design.

Demographic information and laboratory findings were collected on admission, including gender, age, height, weight, blood routine, liver and kidney function tests, electrolytes, and coagulation function. We calculated the prognostic scores MELD and MELD-Na for every patient by using the values of the measured parameters of interest.

Diagnosis of cirrhosis was based on clinical, biochemical, radiological (including ultrasound, elastography, CT and MR) and endoscopic findings related to cirrhosis and/or portal hypertension, or on liver biopsy findings (18). For all registered patients, physicians assessed information on cirrhosis at baseline and 360 days after enrollment.

The area of the skeletal muscle (including psoas major, erector spinae, transverse abdominis, internal and external obliques and rectus abdominis) at the L3 level was measured using the 3D Slicer software (version 5.1.0). SMI was calculated by dividing the area of the skeletal muscle at the L3 level by the square of the height (m²) (13, 14). Analysis of the abdominal CT scans was performed independently by two imaging physicians. A third physician was involved in reaching a consensus whenever disagreement arose.

Statistical analyses were conducted using IBM SPSS Statistics (version 22.0) and R software (version 4.1.2). Continuous variables were expressed as the mean ± standard deviation or median (interquartile spacing) of a normal or non-normal distribution. Categorical variables were shown as numbers (percentages). Comparisons between the two groups were calculated using unpaired t-test for parametric data, Mann-Whitney U test for non-parametric data and chi-square test for categorical data. Univariate and multivariate (forward-stepwise method) logistic regression analyses were conducted to examine potential predictors of cirrhosis. Bilateral P values < 0.05 considered statistically significant.

The receiver operating characteristic curves (ROC) were used to assess the ability of SMI levels to predict the incidence of cirrhosis. Moreover, we constructed a restricted cubic spline (RCS) describing the relationship between SMI and the risk of cirrhosis, according to the three knots of the SMI level distribution, i.e., the 10th, 50th, and 90th percentiles.

A total of 230 male ACLF patients without cirrhosis at baseline were included in this study (Figure 1). The baseline characteristics of the individuals involved are shown in Table 1. During the 360-days follow-up period, 45.2% (104/230) of patients progressed to cirrhosis. Overall, the cirrhosis group was associated with higher mean age, white blood cell count (WBC), international normalized ratio (INR), total bilirubin (TB), MELD score, and MELD-Na score, but lower albumin and blood sodium levels. Patients who progressed to cirrhosis at 1 year presented with a lower SMI [46.1 ± 6.9 versus 49.2 ± 6.5 cm²/m², P = 0.001] and a higher proportion of sarcopenia (19.2% vs. 6.3%, P = 0.003). In addition, patients in the cirrhosis group suffered a greater incidence of hepatic encephalopathy and ascites at baseline (P < 0.05).

To explore the potential association between SMI and long-term incidence of cirrhosis in male non-cirrhotic ACLF patients, univariate and multivariate logistic regression analyses were performed. Univariate logistic regression analyses revealed that age, WBC, INR, and TB were risk factors for long-term cirrhosis onset, while SMI, albumin and blood sodium served as protective factors (Table 2). Next, the above factors were included in a multivariate regression analysis. We found that after adjusting for other factors, high levels of SMI remained a protective agent for 360-days progression to cirrhosis in men with ACLF (OR 0.950, 95% CI: 0.908–0.994, P < 0.05) (Table 2 and Supplementary Figure 1). The RCS analysis demonstrated an approximately linear dose-dependent effect of SMI on the risk of cirrhosis (P for non-linear = 0.127) (Figure 2). The area under the ROC curve (AUC) for the L3-SMI to predict the incidence of cirrhosis in male non-cirrhotic ACLF patients was 0.636 (P < 0.001) (Supplementary Figure 2).

It has been established that skeletal muscle mass is gradually reduced over age (19). Our previous study revealed that muscle loss is particularly pronounced after the age of 40 (13). Besides, we found that L3-PMI was independently associated with 1-year mortality in male ACLF patients younger than 40 years, which was not observed in male ACLF patients beyond 40 years (20). Thus, we assessed SMI levels in the age ≤ 40 years and > 40 years groups, respectively. The mean values of SMI for the older and younger groups were 47.0 ± 6.8 and 48.8 ± 6.8 cm²/m², respectively (P < 0.05). The distribution of SMI in the two groups was shown in Figure 3, exhibiting statistical differences (P < 0.05).

Since there were significant differences in SMI among males with ACLF in different age groups, we further explored the connection between SMI and long-term progression to cirrhosis in various age groups. Of 103 male subjects younger than 40 years, 32 (31.1%) suffered liver cirrhosis after 1 year. In this group, the mean SMI of patients who developed cirrhosis at 1 year was 45.8 ± 6.3 cm²/m², significantly lower than that of non-cirrhotic patients (50.2 ± 6.6 cm²/m², P = 0.003).

We conducted univariate and multivariate logistic regression analyses of patients’ clinical characteristics. The results indicated that higher INR and TB as well as lower SMI were significantly associated with the risk of 360-days cirrhosis in young men no more than 40 years old with ACLF (Table 3). Multivariate logistic regression identified SMI as a protective factor for the long-term progress to cirrhosis (OR 0.908, 95% CI: 0.842–0.979, P < 0.05) (Table 3 and Supplementary Figure 3). Further RCS models revealed an approximately linear dose-response relationship between SMI and the risk of cirrhosis (P for non-linear = 0.905) (Figure 4). The area under the ROC curve (AUC) for the L3-SMI to predict the incidence of cirrhosis in young men (age ≤ 40 years) with ACLF was 0.673 (P < 0.01) (Supplementary Figure 4).

The proportion of male ACLF patients over 40 who developed cirrhosis after 1 year was 56.7% (72/127). The mean SMI of patients with and without cirrhosis after 1 year were 48.0 ± 6.1 cm²/m² and 46.3 ± 7.1 cm²/m², respectively, not showing a statistical difference (P = 0.150). There was no association between SMI and the occurrence of 360-days cirrhosis in older (age > 40 years) male ACLF patients in both univariate and multivariate regression analyses (Table 3). The results of multivariate regression analysis declared that INR and serum sodium were independent predictive factors for the development of 360-days cirrhosis in older men with ACLF (Supplementary Figure 5).