In total, 62 articles were initially identified from the literature search, but only 23 studies fulfilled the inclusion criteria and underwent further evaluation (Suppl. Figure S1) [11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33]. Two studies from a single center in the Netherlands [23, 28] as well as three studies from a single center in China (17,31,33) had overlapping study periods, and therefore only the most recent studies [17, 23] were included. Similarly, three studies from a single center in Canada [12, 29, 30] had overlapping study periods, but the oldest one [12] which provided additional data on patients with myosteatosis was included. Finally, four studies [11, 18, 20, 32] were from the same single center in Italy but two [18, 32] of them had overlapping study periods and we included the newest study [18]. Thus, 17 studies [11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27], that evaluated the prevalence of myosteatosis in cirrhotic patients, fulfilled all inclusion criteria and were included in the final analysis. Four studies were derived from Italy [11, 18, 20, 22], three from China [17, 19, 21], two from Japan [14, 15], as well as two from Germany [16, 25] and the USA [24, 27] and one from Canada [12], Greece [13], South Africa [26] and the Netherlands [23]. In only one study [24] MRI was used for the evaluation of myosteatosis. Ten of the 17 studies had a retrospective design [11, 14,15,16,17, 19, 22,23,24, 26]. The Newcastle–Ottawa scale (NOS), was used to assess the quality of the included studies [34]. Based on that, the studies had low risk of bias (NOS scored > 5) (Suppl. Table 1).

In total, 106 studies were initially identified regarding myosteatosis in patients with chronic liver disease, but only 6 studies met the inclusion criteria [35,36,37,38,39,40] (Suppl. Figure S1). Two studies [38, 40] from a single center in Japan had overlapping study periods, and we included the study with the largest cohort [38]. Thus, 5 studies were finally included [35,36,37,38,39]. In 4 of them [36,37,38,39], the presence of cirrhosis was not an exclusion criterion, but the proportion of patients with cirrhosis was relatively small whenever this was available (Suppl. Table 2). Nevertheless, no separate data were provided for patients with or without cirrhosis in these studies. The included studies were derived from different countries (Korea [35], Taiwan [36], Sweden [37], Japan [38] and Germany [39]), while different definitions were used for myosteatosis (Suppl. Table 2).

In total, 4136 cirrhotic patients [mean age: 60.2 years, 64.5% (2674/4136) males] were evaluated. In the majority of patients (56.3% or 2327/4136) the diagnosis of myosteatosis was based on muscle/m2 radiodensity at L3 < 41 HU for patients with dry body mass index (BMI) < 25 kg/m2 and < 33 HU for those with ΒΜΙ ≥ 25 kg/m2 (i.e. BMI-based definition) [11,12,13,14,15, 18,19,20, 25,26,27]. In 1368 (33%) patients the gender-based definition using different cut offs between males and females (e.g. < 26.6 HU in females and < 28.6 HU in males) [16, 17, 21, 22] was used for diagnosis of myosteatosis, while in 441 (10.7%) patients various other criteria were applied [23, 24]. According to the available data, mean CP and MELD scores were 7.2 and 13.8, respectively. Chronic viral hepatitis (B or C) was the underlying cause of cirrhosis in 39% (n = 1615) of patients, while HCC was present in 39.7% (870/2191) of patients [11, 12, 14, 16, 22, 25,26,27] and 28.8% of the patients (668/2316) had a previous history of hepatic encephalopathy [11,