Significant Correlation Between Grip Strength and m2bpgi in Patients with Chronic Liver Diseases

We sought to compare the impact upon grip strength (GS) between the Mac-2 binding protein glycosylation isomer (M2BPGi) and the Fibrosis-4 (FIB4) index in chronic liver disease (CLD) patients (n = 376: 171 males and 205 females, and 137 liver cirrhosis (LC) cases (36.4%)). Factors linked to the low GS (<26 kg in male and <18 kg in female) were also investigated using univariate and multivariate analyses. The median GS in males was 35.5 kg, while that in females was 21.1 kg. The median M2BPGi was 1.11 cutoff index, whereas the median FIB4 index was 2.069. In both male (P < 0.0001) and female (P = 0.0001), GS in LC patients was significantly lower than that in non-LC patients. In males, M2BPGi (r = −0.4611, P < 0.0001) and the FIB4 index (r = −0.4556, P < 0.0001) significantly correlated with GS. Similarly, in females, M2BPGi (r = −0.33326, P < 0.0001) and our FIB4 index (r = −0.26388, P = 0.0001) also significantly correlated with GS. In the multivariate analyses of factors linked to the low GS, independent factors were: M2BPGi (P = 0.0003) and skeletal muscle index (P = 0.0007) in males, and age (P < 0.0001) and serum albumin level (P = 0.0484) in females. In conclusion, liver fibrosis markers were well-correlated with GS in CLD patients. In particular, M2BPGi can be helpful for predicting the low GS in male patients.


Introduction
Grip strength (GS) is a parameter of muscle function, and muscle dysfunction, such as decreased GS, has been suggested to be a key mechanism of fatigue in chronic liver diseases (CLDs) [1,2]. GS is also an easy and reproducible indication with respect to frailty measures [3]. The clinical implication of GS in patients with CLDs has been verified in several studies. In our previous investigation, we reported that GS had a strong influence especially in the physiological domains in the 36-Item Short-Form Health Survey in patients with CLDs [4]. We also demonstrated that muscle strength decline was closely associated with sleep disorder in patients with CLDs [5], while non-alcoholic fatty liver disease (NAFLD) is reported to be linked to muscular impairment in elderly adults [6][7][8]. Lee et al. reported that there were linear decreases in the NAFLD index across incremental GS values [8]. A previous large, multicenter, longitudinal population study with varying incomes and sociocultural backgrounds, showed that the measurement of GS is a simple, inexpensive risk-stratifying screening tool for all-cause death, cardiovascular-related death, and cardiovascular disease [9]. Thus, factors relevant to decreased GS are of importance in the clinical settings.
A novel liver fibrosis marker (Mac-2 binding protein glycosylation isomer (M2BPGi)), which is a glycobiomarker associated with CLD-related liver fibrosis with a unique fibrosis-related glycoalteration, has been recently established by Japanese researchers [10][11][12][13]. The usefulness of M2BPGi for the prediction of the severity of liver fibrosis or liver carcinogenesis has been well confirmed, while the FIB4 index is also a well-validated liver fibrosis marker in patients with CLDs [13][14][15][16][17][18][19][20][21]. A recent large observational study reported that patients with liver cirrhosis (LC) before a sustained virological response (SVR) to treatment for the hepatitis C virus (HCV) infection continue to have a significant risk for hepatocellular carcinoma (HCC) development (over 2% per year) for a large number of years, even when their FIB4 index decreases, and should continue HCC surveillance, while non-LC patients with FIB4 index 3.25 or more have a higher risk to merit HCC surveillance, especially when FIB4 index remains 3.25 or more post-SVR [19].
However, there have been no reports examining the impact of M2BPGi or FIB4 index on GS in CLD patients. In this study, we sought to compare the impact on GS between M2BPGi and FIB4 index in CLD patients.

Patients
A total of 376 chronic liver disease (CLD) individuals with data for both grip strength (GS) and Mac-2 binding protein glycosylation isomer (M2BPGi) were admitted to our hospital between February 2014 and January 2019, and they were subjected to this analysis. GS was tested based upon the current Japanese guidelines [22]. For the measurement of GS, we used a Smedley grip dynamometer. The dynamometer was placed so that the needle was on the lateral side, and the patient stood upright with both of their legs spread naturally, and both arms hanging down naturally. The measurement was done in this state by having the patient firmly grip the device while making sure it did not come into contact with the body or with the clothing. Two measurements were done on both sides. The better measurement on each side was chosen, and GS was calculated as the average of these values [22]. Liver cirrhosis (LC) was determined based on pathological data, radiologic findings and/or laboratory data [23][24][25][26]. HCC diagnosis was based on previous reports [27]. M2BPGi was measured as reported elsewhere [28]. The fibrosis-4 (FIB4) index was calculated as reported previously [29]. The skeletal muscle index (SMI, kg/m 2 ) was measured using bioimpedance analysis (BIA), as reported elsewhere [30]. Patients with severe ascites were excluded because of the lack of reliability for BIA testing. We retrospectively examined the relationship between the baseline GS value and clinical parameters. The low GS was defined as <26 kg in male and <18 kg in female based on the current guidelines [22,31]. Factors linked to the low GS were also investigated using univariate and multivariate analyses.
This study protocol was acknowledged by the institutional review board in the Hyogo College of Medicine Hospital (approval no. 1831) and the Declaration of Helsinki was strictly followed in order to guarantee the right of the study subjects. All patients provided written informed consent. Personal information was protected during data collection.

Statistical Considerations
In the analysis of continuous parameters, we employed Student's t test, the Mann-Whitney U test, Pearson's correlation coefficient r, analysis of variance or Kruskal-Wallis test to assess group difference, as appropriate. In the analysis of categorical parameters, we employed a Pearson χ 2 test to assess group difference. Baseline items significantly correlated with the low GS in our univariate analysis were also entered into the multivariate logistic regression analysis to select candidate items. Unless otherwise stated, data were presented as a median value (interquartile range (IQR)). We set the threshold for statistical significance at P < 0.05. The JMP 14 (SAS Institute Inc., Cary, NC, USA) was employed to analyze statistically.

GS According to Age in Male and Female
Patients were classified into three groups according to age: <50 years (group A); ≥50 years and <65 years (group B); and ≥65 years (group C). In males, the median (IQR) GS in groups of A (n = 72), B (n = 46) and C (n = 53) were: 39.

GS According to Body Mass Index (BMI) in Male and Female
Patients were classified into three groups according to BMI: ≥25 kg/m 2 (Group A); ≥20 kg/m 2 and <25 kg/m 2 (Group B); and <20 kg/m 2 (Group C). In males, the median (IQR) GS in groups of a (n = 24), b (n = 94) and c (n = 53) were: 37.

GS According to Age in Male and Female
Patients were classified into three groups according to age: <50 years (group A); ≥50 years and <65 years (group B); and ≥65 years (group C). In males, the median (IQR) GS in groups of A (n = 72), B (n = 46) and C (n = 53) were: 39.

GS According to Age in Male and Female
Patients were classified into three groups according to age: <50 years (group A); ≥50 years and <65 years (group B); and ≥65 years (group C). In males, the median (IQR) GS in groups of A (n = 72), B (n = 46) and C (n = 53) were: 39.

GS According to Body Mass Index (BMI) in Male and Female
Patients were classified into three groups according to BMI: ≥25 kg/m 2 (Group A); ≥20 kg/m 2 and <25 kg/m 2 (Group B); and <20 kg/m 2 (Group C). In males, the median (IQR) GS in groups of a (n = 24), b (n = 94) and c (n = 53) were: 37.

GS According to Body Mass Index (BMI) in Male and Female
Patients were classified into three groups according to BMI: ≥25 kg/m 2 (Group A); ≥20 kg/m 2 and <25 kg/m 2 (Group B); and <20 kg/m 2 (Group C). In males, the median (IQR) GS in groups of a (n =

Correlation Between GS and m2bpgi in Male and Female
In

Correlation Between GS and Fib-4 Index in Male and Female
In males, the FIB-4 index significantly correlated with GS for all cases (r = −0.4556, P < 0.

Uni-and Multivariate Analyses of Factors Lnked to Low GS in Female Patients
In females, the univariate analysis observed seven factors to be significantly associated with low GS (<18 kg): Age (P < 0.0001), serum albumin level (P = 0.0209), PT (P = 0.0283), platelet count (P = 0.0255), M2BPGi (P = 0.0017), FIB-4 index (P = 0.0038) and SMI (P = 0.0036). (Table 4) Multivariate analysis for the seven factors showed that age (P < 0.0001) and serum albumin level (P = 0.0484) were significant factors linked to the low GS. (Table 5) HRs and 95% CIs for these items were indicated in Table 5.

Discussion
Liver and muscle are both metabolically active endocrine organs, and CLDs and sarcopenia, as defined by muscle mass loss and muscle strength decline, may share common pathogenic determinants [32,33]. However, to our knowledge, there have been few studies of factors relevant to muscle strength decline, especially focusing on liver fibrosis markers. There is a high correlation between GS and other muscle strength, and thus GS can be considered as an index of muscle strength throughout the body. Reduced muscle strength, as defined by decreased GS, has been associated with an elevated risk of mortality in many preceding researches [9,[34][35][36]. Thus, to elucidate factors associated with GS decline may be clinically meaningful.
In our data, the median GS values were 35.5 kg in males and 21.1 kg in females. The average GS values in Japanese adults in their 60s are reported to be around 40 kg in males and around 25 kg in females, which are higher than our data [37]. CLD itself may cause these results. In our results, M2BPGi had significant negative correlation with GS irrespective of the LC status or gender, while in the FIB-4 index, similar tendencies were observed except for female LC cases. Our LC patients had significantly low GS compared with non-LC patients, which can explain the significant relationship between GS and liver fibrosis markers. M2BPGi can be a surrogate marker for assessing hepatic stellate cells (HSCs) activation, which are associated with liver fibrosis progression [13]. HSCs are also reported to be involved in extrahepatic disease progression [13]. In CLD patients with low GS, close monitoring for liver fibrosis may be required. While the FIB-4 index includes age [29], as shown in Figure 2A,B, GS decreases with increasing age both in males and females, which can be linked to our current results.
In our multivariate analysis for male patients, M2BPGi (P = 0.0003) and SMI (P = 0.0007) were significant factors linked to the low GS. In male CLD patients, M2BPGi rather than the FIB-4 index may affect GS. Sung et al. reported that hepatic encephalopathy, higher M2BPGi, higher age and lower GS were independent predictors of skeletal muscle decline [38]. Our results were consistent with their results. While in our multivariate analysis for female patients, age (P < 0.0001) and serum albumin level (P = 0.0484) were independent predictors for the low GS. Although the reasons for these discrepancies between genders in the multivariate analyses are unclear, clinicians should be aware that factors influencing on GS can be different between genders. On the other hand, contrary to our expectations, the serum ammonia level was not associated with low GS both in males and females in the univariate analyses. We have previously reported that hyperammonemia in LC patients correlated with a higher serum myostatin level, which strongly suppresses skeletal muscle growth [39]. In our hypothesis, hyperammonemia caused hypermyostatinemia and subsequent muscle strength decline. Further investigations will be needed to confirm the impact of hyperammonemia on GS decline.
SMI had a significant correlation with GS both in male (r = 0.451957, P < 0.0001) and female (r = 0.427142, P < 0.0001) in this study, which suggests the close linkage between muscle strength and muscle mass. In the current guidelines for sarcopenia, muscle function is assessed firstly and in patients with decreased muscle strength, muscle mass is assessed secondly [22,31,40].
Thus, clinical implications of decreased muscle mass with preserved muscle strength should be clarified in future studies. On the other hand, in NAFLD patients, GS is shown to correlate with steatosis grade [6][7][8]. Although investigation of the impact of GS on steatosis in NAFLD patients is beyond the scope of our analysis, to confirm these in Japanese NAFLD patients may be essential.
Several limitations related to the study warrant mention. First, the study was a single-center observational study with a retrospective nature. Second, the study data was derived from a Japanese liver disease population data, and additional investigations on other races are required to further verify and extend the application to other races. Third, GS can vary depending on patients' daily life activities. Fourth, patients with massive ascites who are potentially involved in the low GS were excluded due to the lack of reliability in the BIA, creating bias. Finally, the interpretation of our results should be done cautiously, since the direction of the association between baseline data and GS remains unclear, due to the cross-sectional nature of our data. Nevertheless, our study results denoted that liver fibrosis markers, especially M2BPGi, can be closely associated with the GS decline.

Conclusions
In conclusion, M2BPGi can be a useful marker for the screening of patients with GS decline.