The systematic review was conducted according to the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines (19). The protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) (registration number: CRD42023441039).

According to the definition of the European Working Group of Sarcopenia in Older People (EWGSOP2) (2), sarcopenia is considered a progressive and generalized skeletal muscle disorder characterized by the gradual loss of muscle strength, mass and function.

Based on previous studies, we included studies that reported sarcopenia of any severity under any definition criteria.

Following the PICOS question, we included:

We excluded the following studies: conference papers, case reports, letters to the Editor, reviews, studies that did not provide sufficient data, non-Chinese and non-English literature.

Several relevant databases were comprehensively searched for this systematic review, including PubMed, Web of Science, Embase, the Cochrane Library, the China Biomedical Literature Service System, China Biomedical Literature Service System (CINAHL), China National Knowledge Infrastructure (CNKI), the WeiPu (VIP) and the Wanfang database, from inception to 1 August 2024.

The following search was used in PubMed:

((((liver failure[Title/Abstract]) OR (hepatic failure[Title/Abstract])) OR (acute-on-chronic liver failure[Title/Abstract])) OR (ACLF[Title/Abstract])) OR (end-stage liver disease[Title/Abstract]) OR ((((liver failure[MeSH Terms]) OR (hepatic failure[MeSH Terms])) OR (acute-on-chronic liver failure[MeSH Terms])) OR (end-stage liver disease[MeSH Terms]))

AND

(((((((((((sarcopenia[Title/Abstract]) OR (sarcopenic[Title/Abstract])) OR (muscle loss[Title/Abstract])) OR (muscular atrophy[Title/Abstract])) OR (less muscle disease[Title/Abstract])) OR (muscle wasting[Title/Abstract])) OR (muscle mass[Title/Abstract])) OR (muscle weakness[Title/Abstract])) OR (muscle strength[Title/Abstract])) OR (hand strength[Title/Abstract])) OR (grip strength[Title/Abstract]) OR ((((((((sarcopenia[MeSH Terms]) OR (muscular atrophy[MeSH Terms])) OR (less muscle disease[MeSH Terms])) OR (muscle wasting[MeSH Terms])) OR (muscle mass[MeSH Terms])) OR (muscle weakness[MeSH Terms])) OR (muscle strength[MeSH Terms])) OR (hand strength[MeSH Terms])) OR (grip strength[MeSH Terms])).

Two researchers screened the articles and extracted data independently, utilizing EndNote software. Disagreements between researchers were resolved through discussion or consultation with a third reviewer in necessity. Data is extracted from eligible articles using a standardized form. The form included: the first author, publication year, types of study, age, regions, observation period, sample size, percentage of women, definitions and measurements of sarcopenia, ACLF definition, outcomes, follow-up period, reported HR (95%CI), and adjusted variables.

Two researchers independently evaluated the quality of cohort studies by the Newcastle-Ottawa Scale (NOS) (20). A discussion was made if there was a disagreement. The scale covers three domains and eight items, including the selection of research subjects (4 items, 4 points), comparability between groups (1 item, 2 points), and outcome measurement (3 items, 3 points). The total score for every study was 9. The study was considered moderate to high quality if the score was more than six.

The data analysis was conducted using RevMan 5.4 and Stata 12.0. Meta-analysis was performed using a random-effects model to generate a pooled hazard ratio (HR) and 95% confidence interval (CI). When I² ≥ 50% or p < 0.05, heterogeneity was determined to be statistically significant. To explore the source of heterogeneity in the study, we performed a series of subgroup analyses. In addition, to evaluate publication bias by observing the symmetry of the funnel plot, Begg’s/ Egger’s tests and combined with trim and fill method. Sensitivity analysis was performed by excluding each study to evaluate the stability of the results. For all statistical tests, the test level of meta-analysis was set at α = 0.05.

Figure 1 summarizes the yield of the search process. A total of 4,014 records were identified from the database, including 567 duplicates. Following the removal of duplicates, 3,187 records passed through title and abstract screening, and 35 articles were retained for full-text review. Ultimately, 12 total studies were included in the qualitative and quantitative syntheses, collectively representing approximately 2,505 participants. Of these, 11 studies were identified for meta-analysis and included approximately 2,072 participants. The study of Peng et al. (21) was excluded from the meta-analysis, as its outcomes significantly differed from those reported by other 11 studies.

Table 1 reports the main characteristics of the included studies in the systematic review.

The Newcastle–Ottawa Scale (20) was used to assess the risk of bias for the 12 studies included in the systematic analysis, with the results presented in Table 2. Two studies scored 8 or 9 points, indicating a low risk of bias. Ten studies scored 6 or 7 points, indicating a medium risk of bias. Consequently, since all cohort studies achieved a score of 6 or higher, no studies were excluded from the systematic review.

Subgroup analyses showed that the mortality risk was significantly increased in the study subgroup with short-term mortality outcomes [HR = 1.54, 95%CI (1.14, 2.08)] and prospective study [HR = 2.16, 95%CI (1.30, 3.60)]. Other subgroup analyses did not show a significant association between sarcopenia and the mortality risk.

By conducting subgroup analyses stratified by outcomes, regions, methods for measuring sarcopenia, and sample size, we did not find a significant reduction in heterogeneity, indicating that heterogeneity was not associated with these variables. Heterogeneity was significantly reduced in subgroup analyses of research type, suggesting that heterogeneity might have originated from research type. This may be related to the potential recall bias inherent in retrospective cohort studies compared to prospective cohort studies. Furthermore, the high heterogeneity may also be related to variations in the original studies, including non-standardized definitions of sarcopenia, differences in baseline disease severity among patients, and other potential comorbidities (malnutrition/ inflammation) (Table 3).

In the 12 included studies, only 2 cohort studies (28, 29) reported infection-related outcomes.

Of these, the research of Wang et al. (28) suggested that low L3-SMI is an independent risk factor for the development of infections and significantly influences [OR = 0.89, 95%CI (0.81–0.97), p = 0.011]. However, Zhou, et al. (29) found no significant difference in the incidence of in-hospital infections between patients with low ESI and those with ACLF [HR = 1.62, 95%CI (0.6–3.08), p = 0.138]. Therefore, the association between sarcopenia and infection in patients with ACLF remains unclear. Further research is necessary to elucidate this association.

The systematic review and meta-analysis reviewed data from 12 cohort studies and analyzed the effect of sarcopenia on the prognosis of patients with ACLF. Our meta-analysis found that sarcopenia is significantly associated with an increased mortality risk in patients with ACLF. However, we did not find a significant association between sarcopenia and the risks of infection in patients with ACLF.

In recent years, an increasing number of studies have been conducted to explore the association between sarcopenia and other diseases, such as chronic liver disease (30), neoplastic diseases (31), and liver transplantation (32), et al. Sarcopenia is closely associated with the prognosis of these diseases. However, few studies have specifically examined the effect of sarcopenia on patients with ACLF. Currently, the mechanisms through which sarcopenia influences the prognosis of patients with ACLF remain unclear.

The mechanisms by which sarcopenia influences the prognosis of patients with ACLF may be attributable to systemic inflammation, and hyperammonemia, et al. First, systemic inflammation is the key contributor to the loss of skeletal muscle and its functionality (33). Also, Systemic inflammation is one of the features of ACLF and is intimately associated with the prognosis of patients (34). For patients with ACLF, various risk factors can lead to acute hepatic decompensation, which in turn causes systemic inflammation based on chronic liver disease. ACLF often manifests as acute systemic inflammatory response syndrome (35). Of these, pro-inflammatory factors [tumor necrosis factor (TNF-α), interleukin (IL-1β), and interferon (IFN-γ)] (33) can inhibit the expression of growth hormone and promote the occurrence of sarcopenia. Meanwhile, pro-inflammatory factors also activate the ubiquitin-proteasome pathway, causing protein degradation, leading to sarcopenia (36). When patients with ACLF are complicated by sarcopenia, their bodies are still unable to cope with the catabolic changes, even if liver function is compensated (37). Therefore, sarcopenia may be a manifestation of potential systemic inflammation, which could influence the prognosis of patients with ACLF. Furthermore, Ye et al. (38) conducted a cohort study of 140 patients with ACLF and suggested that white blood cells (WBC) were significantly elevated in patients with sarcopenia, which indirectly confirmed the potential relationship among sarcopenia-inflammation-ACLF. Second, as a common complication of patients with ACLF, hyperammonemia may reduce muscle protein synthesis by upregulating myostatin production (39). Finally, patients with ACLF often experience a reduced oral intake due to complications of the portal system and impaired liver function (40). This diminished dietary protein intake is also among the contributing factors to the onset of sarcopenia. Also, systemic inflammation and hyperammonaemia can contribute to a heightened catabolic state and collectively lead to malnutrition and sarcopenia (16).

As we all know, patients with ACLF have a very high short-term mortality when treated without liver transplantation (41). Our study found that the short-term mortality risk (≤90 days mortality) [HR = 1.54, (1.14, 2.08)] is higher than the total mortality risk [HR = 1.27, 95%CI (1.05, 1.54)] of patients with ACLF who are complicated by sarcopenia. Therefore, sarcopenia may be a reliable indicator to predict the prognosis of patients with ACLF in the future. The death of patients with ACLF is associated with multi-organ failure. Sarcopenia reduces the body’s compensatory capacity, making patients more susceptible to severe complications such as hepatorenal syndrome and hepatic encephalopathy, thereby significantly increasing 28-day and 90-day mortality rates. Studies suggested that the mortality of ACLF increased with the increasing incidence of organ failure. Patients with one organ failure have a mortality rate of about 20% within 28 days, while those with three organ failures have a mortality rate of more than 70% (42). Liver transplantation has been demonstrated to improve the survival rates of patients with ACLF significantly, but the applicability and feasibility of liver transplantation are limited by various factors.

Infection was deemed to be a major cause of increased mortality in patients with ACLF (43). Identification of infection-related risk factors in ACLF patients could facilitate the development of multidisciplinary therapeutic strategies, thereby potentially improving clinical outcomes and reducing mortality. A recent study shows that sarcopenia is a highly predictive nutritional indicator for the occurrence of hospital-acquired infections (44). The study of Wang et al. (28) suggested that sarcopenia is an independent risk factor for infections in ACLF patients. The elevated infection risk associated with sarcopenia may be mediated through immune system compromise, as muscle wasting has been shown to impair both innate and adaptive immune responses in chronic disease states (45). However, the study of Zhou et al. (29) found no significant difference in the incidence of in-hospital infections between patients with low ESI and those with ACLF. Differences across studies stems from multiple sources, including the operational definition and measurement modality used for sarcopenia, baseline demographic and clinical characteristics of the enrolled populations, sample size, and other methodological factors. Thus, the causal relationship between sarcopenia and infection—whether sarcopenia predisposes to infections or infections accelerate muscle wasting—remains to be elucidated through prospective studies. Clarifying this bidirectional association could inform novel therapeutic and preventive strategies.

The current meta-analysis has the following strengths. First, according to our knowledge, our research is the first meta-analysis to evaluate the effects of sarcopenia on the prognosis of patients with ACLF. Second, our study strictly adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) 2020 guidelines. Finally, we used several methods to fully test the stability of the results, encompassing sensitivity analyses and subgroup analyses.

Besides, our research still presents some potential limitations. First, these findings should be interpreted with caution due to the limited number of included studies and higher heterogeneity in the analysis. Besides, we conducted further analysis to evaluate the source of this heterogeneity. The subgroup analysis found that interstudy heterogeneity may be associated with research type. Second, we cannot extract the HR from the study about sarcopenia and infection in ACLF patients. Therefore, we only conducted a descriptive analysis to explore the association between sarcopenia and infection in ACLF patients. Third, there may be potential publication bias in our meta-analysis. Finally, the majority of the studies we included were retrospective cohort studies, which means we cannot infer the causal relationship between sarcopenia and the mortality of patients with ACLF. We recommend that future high-quality prospective studies be conducted to validate our findings.