Lower muscular strength is associated with greater liver fat content and higher serum liver enzymes—“The Sedentary's Liver” The Study of Health in Pomerania

Abstract

We investigated the associations of low handgrip strength (HGS, i.e., a marker of muscular fitness) with liver fat content (LFC) and serum liver enzymes in a population‐based setting. We used data from 2700 participants (51.7% women), aged 21–90 years, from two independent cohorts of the population‐based Study of Health in Pomerania (SHIP‐START‐2 and SHIP‐TREND‐0). Cross‐sectional, multivariable adjusted regression models were performed to examine the associations of HGS with LFC, measured by magnetic resonance imaging and serum liver enzymes. We found significant inverse associations of HGS with both LFC and serum liver enzymes. Specifically, a 10‐kg lower HGS was associated with a 0.59% (95% confidence interval [CI]: 0.24–0.94; p = 0.001) higher LFC, a 0.051 µkatal/L (95% CI: 0.005–0.097; p = 0.031) higher gamma‐glutamyltransferase (GGT) concentration and a 0.010 µkatal/L (95% CI: 0.001–0.020; p = 0.023) higher aspartate aminotransferase (AST) concentration. The adjusted odds‐ratio for prevalent hepatic steatosis (defined by a MRI‐PDFF ≥5.1%) per 10‐kg lower HGS was 1.21 (95% CI: 1.04–1.40; p = 0.014). When considering only obese individuals, those with low HGS had a 1.58% (95% CI: 0.18–2.98; p = 0.027) higher mean LFC and higher chance of prevalent hepatic steatosis (adjusted OR 1.74, 95% CI: 1.15–2.62; p = 0.009) compared to individuals with high HGS. We found similar associations in individuals with overweight, but not in those with normal weight. Lower HGS was strongly associated with both higher LFC and higher serum GGT and AST concentrations. Future studies might clarify whether these findings reflect adverse effects of a sedentary lifestyle or aging on the liver.

Highlights

• We used magnetic resonance imaging, the most accurate and sensitive noninvasive diagnostic tool for determination of hepatic steatosis

• Lower handgrip strength was strongly associated with both higher liver fat content and higher serum GGT and AST concentrations

• Especially overweight and obese individuals with low handgrip strength had a significant higher risk of prevalent hepatic steatosis

Aging‐related decrease of muscular fitness and its associations with liver parameters.

Abbreviations

ALT

alanine aminotransferase

AST

aspartate aminotransferase

BMI

body mass index

CI

confidence interval

GGT

gamma‐glutamyltransferase

HGS

handgrip strength

LFC

liver fat content

MRI

magnetic resonance imaging

NAFLD

non‐alcoholic fatty liver disease

OR

odds ratio

PDFF

proton density fat fraction

SHIP

Study of Health in Pomerania

T2D

type 2 diabetes

1. INTRODUCTION

In Western societies, nonalcoholic fatty liver disease (NAFLD) is the most common chronic liver disease (Byrne et al., 2015). NAFLD ranges from hepatic steatosis to nonalcoholic steatohepatitis, with or without coexisting fibrosis and cirrhosis and increases the risk for hepatocellular carcinoma (Byrne et al., 2015; Han et al., 2022). The global prevalence of NAFLD in adults is estimated to be around 25%–30% (Byrne et al., 2015) and increases up to ∼70%–90% among subjects who are obese or have type 2 diabetes (T2D) (Byrne et al., 2015; Younossi et al., 2016). Growing evidence shows that the presence of NAFLD not only increases liver‐related morbidity and mortality, but is also associated with higher risk of developing other cardiometabolic diseases, such as cardiovascular disease, chronic kidney disease, or T2D (Byrne et al., 2015).

Sedentary behavior became the predominant lifestyle in Europe with an average sedentary time of 8–9 h per day (Silva et al., 2020). Many studies have highlighted the deleterious effects of a sedentary lifestyle on cardiovascular health, metabolic diseases such as T2D and NAFLD, as well as on physical fitness (Bowden Davies et al., 2019; Silva et al., 2020).

Physical fitness can be divided into cardiorespiratory and muscular fitness (Markus et al., 2021a). Previous studies (Drzyzga et al., 2021; Markus et al., 2021a, 2021b, 2021c) of our group analyzed associations of cardiorespiratory fitness (Drzyzga et al., 2021; Markus et al., 2021b, 2021c) and handgrip strength (HGS) (Markus et al., 2021a) with the structure and function of the heart. We found that low cardiorespiratory fitness and low HGS were both associated with a smaller and stiffer heart, which we have identified as “the sedentary's heart”. In another cross‐sectional study (Zinterl et al., 2022), we showed that lower cardiorespiratory fitness was associated with greater liver fat content (LFC) and higher serum gamma‐glutamyltransferase (GGT) concentrations: the so‐called “the sedentary's liver”. Muscular fitness consists of muscular endurance, mass, and strength (Markus et al., 2021a). During the aging process, the muscular strength reaches a peak during the second or third decade of life and then follows a slow decline, which is amplified after the age of 65 (Markus et al., 2021a). Muscular strength can be assessed by handgrip strength measurement (Markus et al., 2021a), which correlates well to the whole‐body muscle strength (Carbone et al., 2020). Growing evidence showed that muscular strength is inversely associated with prevalent cardiovascular diseases, overall mortality, and metabolic diseases (Carbone et al., 2020), including NAFLD (Han et al., 2022). However, to our knowledge, none of the studies included in the aforementioned review (Han et al., 2022) used magnetic resonance imaging proton density fat fraction (MRI‐PDFF) to diagnose NAFLD, which is considered a more accurate diagnostic method than ultrasound or hepatic steatosis index (Piazzolla et al., 2020). Therefore, the aim of our study was to analyze the association of HGS with LFC, detected by MRI‐PDFF, and serum liver enzymes in a population‐based cohort of German adults.

2. MATERIALS AND METHODS

2.1. Study population

The data for the present project was taken from the population‐based Study of Health in Pomerania (SHIP). The study design and recruitment strategy have been described in detail previously (Völzke et al., 2022). We performed cross‐sectional analyses using pooled data from SHIP‐START‐2 (2008–2012) and SHIP‐TREND‐0 (2008–2012). The initial cohort included 6753 participants (3510 women [52.0%]). From the 6753 participants, we used a sub‐cohort of 2788 participants with available liver MRI data for measuring LFC for the present study. From this sub‐cohort, we excluded participants with previous self‐reported cirrhosis and hepatitis (n = 19). We also excluded participants with missing values for HGS measurement, serum liver enzyme levels, or any of the considered confounders (n = 69). Thus, the final cohort comprised a total of 2700 participants (1396 women [51.7%]), aged 21–90 years (Supplementary Figure S1).

All participants gave written informed consent. The study was approved by the ethics committee of the University of Greifswald (Völzke et al., 2022) and complies with the Declaration of Helsinki.

2.2. Handgrip strength measurement

The HGS, expressed in kilograms (kg), was measured by a handheld dynamometer (Smedley's Dynamometer, Scandidact, Odder, Denmark) (Markus et al., 2021a). Standing participants were instructed to keep the upper arm close to the trunk and bring the elbow in a 90° flexion. Afterward they gripped the handle with the maximum effort for 3 seconds, one time with each hand. For further analysis the highest value, whether from the right or left hand, was taken (Markus et al., 2021a).

2.3. Bioelectrical impedance analysis

Fat mass and fat‐free mass were measured by bioelectrical impedance analysis using a multifrequency Nutriguard‐M device (Data Input) and the NUTRI4 software (Data Input) (Kohler et al., 2018; Kyle et al., 2004).

2.4. Liver magnetic resonance imaging

Liver MRI was conducted by a 1.5T MRI system (Magnetom Avanto, software version VB15; Siemens Healthineers Erlangen, Germany) with a 12‐channel phased‐array surface coil (Kühn et al., 2017). The assessment of LFC was performed by an experienced radiologist using offline reconstruction of a proton density fat fraction (PDFF) map (Kühn et al., 2017). The radiologist was unaware of the participants' clinical data (Kühn et al., 2017).

The analyzed mean PDFF values were taken from operator‐defined regions of interest at the center of the liver, by using Osirix (v3.8.1; Pixmec Sarl) (Kühn et al., 2017). Presence of hepatic steatosis was defined as MRI‐PDFF ≥5.1%.

2.5. Laboratory data

For laboratory examinations, non‐fasting blood samples were drawn from the cubital vein in all participants (Pitchika et al., 2021). Serum concentrations of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and gamma‐glutamyltransferase (GGT) were measured photometrically using the Dimension Vista 500 analytical system (Siemens Healthcare Diagnostics, Eschborn, Germany) (Pitchika et al., 2021) and were expressed as μkatal/L.

2.6. Statistical analysis

Continuous variables were expressed as medians (25th and 75th percentile) and categorical variables as absolute numbers (percentages), stratified by high and low HGS values. High and low HGS values were defined according to values either above (HGS > median) or below (HGS ≤ median) the fat‐free mass specific median. We standardized HGS to fat‐free mass, which is considered a better predictor for functional decline and outcome when compared to body mass index (BMI) (Merchant et al., 2021). While absolute HGS increases with increasing body weight, the relative HGS (considering body composition) shows an inverse association with body weight (Fogelholm et al., 2006; Lawman et al., 2016). We performed a median regression with fat‐free mass (as independent variable) and HGS as the outcome and derived a HGS cutoff for each specific level of fat‐free mass. This was performed to derive a specific median of HGS for each fat‐free mass value in the population to categorize probands into low (i.e., equal to or below the median) and high (above the median) HGS groups.

For the analyses of the associations of HGS with LFC and serum liver enzyme levels (ALT, AST, and GGT), we performed multivariable linear regression models adjusted for age, sex, fat mass, fat‐free mass, self‐reported T2D, hypertension, smoking status, alcohol consumption, and use of lipid‐lowering medications. For the association of low versus high HGS with LFC and serum liver enzymes in different BMI groups, multivariable linear regression models were also conducted and adjusted for age, sex, self‐reported T2D, hypertension, smoking status, alcohol consumption, and use of lipid‐lowering medications. The odds ratio (OR) for the association between HGS and prevalent hepatic steatosis was calculated by logistic regression models that were adjusted for age, sex, fat mass, fat‐free mass, self‐reported T2D, hypertension, smoking status, alcohol consumption and use of lipid‐lowering medications (when stratified by BMI levels, no adjustment was made for fat mass and fat‐free mass).

A p‐value <0.05 was considered as statistically significant. Statistical analyses were performed using STATA 17.0 (Stata Corporation).

Please see the online data supplement for a more detailed description.

3. RESULTS

3.1. Characteristics of the study population stratified by high and low HGS

The study population comprised 2700 adult participants, 1396 women [51.7%] and 1304 men [48.3%], with a mean age of 52 years. In total, 40% of the study participants had hepatic steatosis (defined as MRI‐PDFF ≥5.1%). When stratified the participants by high and low fat‐free mass‐specific median of HGS the prevalence of hepatic steatosis was significantly greater in participants with low HGS than in those with high HGS (51% vs. 29%). Compared to those with high HGS, participants with low HGS were older (57 vs. 47 years) and more often women (62% vs. 42%). Moreover, compared to those with high HGS, participants with low HGS were more likely to be overweight or obese (81% vs. 59.1%), and had hypertension, T2D, and previous cardiovascular diseases with consequently more prevalent use of antihypertensive, hypoglycemic, or lipid‐lowering medications. In addition, participants with low HGS consumed less amount of alcohol and were more often never smokers (Table 1). For characteristics of the study population stratified by sex and by high and low fat free‐mass specific median of HGS please see Supplementary Table S1.

TABLE 1.

Characteristics of the study population stratified by high and low fat free‐mass specific median ^a of handgrip strength (n = 2700).

Parameter	High HGS	Low HGS	p‐value
N	1338	1362
Handgrip strength (kg)	44 (33; 52)	29 (24; 40)	<0.001
Age (years)	47 (38; 57)	57 (47; 67)	<0.001
Women (%)	41.6	61.7	<0.001
Fat‐free mass (kg)	57.2 (46.5; 65.7)	53.1 (47.1; 66.6)	<0.001
Fat mass (kg)	19.5 (15.6; 23.8)	25.5 (19.5; 31.8)	<0.001
Body mass index (kg/m2)	25.8 (23.4; 28.2)	29.1 (26.0; 32.2)	<0.001
Normal weight (%)	40.9	19.0	<0.001
Overweight (%)	45.3	38.7
Obesity (%)	13.8	42.3
Waist circumference (cm)	86 (78; 94)	94 (83; 104)	<0.001
Systolic blood pressure (mmHg)	125 (114; 136)	129 (117; 141)	<0.001
Diastolic blood pressure (mmHg)	77 (71; 84)	78 (71; 85)	0.134
Hypertension (%)	34.3	56.2	<0.001
Antihypertensive medication (%)	21.2	44.0	<0.001
Glycated hemoglobin (%)	5.2 (4.8; 5.5)	5.3 (5.0; 5.7)	<0.001
Type 2 diabetes (%)	3.7	11.2	<0.001
Hypoglycemic medication (%)	2.2	7.7	<0.001
Total cholesterol (mmol/L)	5.4 (4.7; 6.1)	5.5 (4.7; 6.3)	0.280
LDL cholesterol (mmol/L)	3.3 (2.7; 3.9)	3.3 (2.7; 4.0)	0.439
HDL cholesterol (mmol/L)	1.5 (1.2; 1.7)	1.4 (1.2; 1.7)	0.003
Lipid‐lowering medication (%)	6.7	14.7	<0.001
Estimated glomerular filtration rate (mL/min/1.73 m2)	84 (74; 96)	79 (68; 92)	<0.001
Smoking status (%)			<0.001
Never	35.1	42.7
Former	37.2	38.3
Current	27.7	18.9
Alcohol consumption (g/day)	5.5 (1.5; 13.1)	3.3 (0.7; 9.1)	<0.001
Sedentary lifestyle (%)	27.6	30.2	0.133
Previous cardiovascular disease (%)	1.6	3.5	0.002

Note: Data are medians (25th, 75th percentile) or percentages. p‐values were derived from the Wilcoxon tests (for continuous data) or Chi‐squared tests (for categorical data).

^a The low and high fat‐free mass specific medians of handgrip strength were calculated as a cutoff for each specific level of fat‐free mass.

3.2. Reversion of the x‐axis scale to run from maximum HGS value to minimum

Older age was associated with lower HGS values and higher values of LFC (Supplementary Figure S2). As the aim of our primary analysis was to investigate the differences of LFC and serum liver enzyme levels with lowering HGS values (after adjustment for age and other potential confounders), we reported all figures with a reversed x‐axis, running from the highest HGS value to the lowest one, thus permitting a more intuitive interpretation of our results.

3.3. Associations of HGS values with LFC and serum liver enzymes

As a result of our multivariable‐adjusted regression analyses, we found significant inverse associations of HGS with LFC and serum liver enzymes, mainly serum AST and GGT levels (Figure 1A,C,D). Specifically, a 10‐kg lower HGS was associated with a 0.59% (95% confidence interval [CI]: 0.24–0.94; p = 0.001) higher LFC, a 0.010 µkatal/L (95% CI: 0.001–0.020; p = 0.023) higher serum AST concentration and a 0.051 µkatal/L (95% CI: 0.005–0.097; p = 0.031) higher serum GGT concentration. No significant association was found between HGS and serum ALT level (Figure 1B).

FIGURE 1.

Adjusted^# regression line (95% CI) showing the association of handgrip strength (HGS) with (A) liver fat content (LFC), (B) alanine aminotransferase (ALT), (C) aspartate aminotransferase (AST), and (D) gamma‐glutamyltransferase (GGT) (n = 2700). ^#Linear regression adjusted for age, sex, fat mass, fat‐free mass, self‐reported type 2 diabetes, hypertension, smoking status, alcohol consumption, and use of lipid‐lowering medications.

The multivariable adjusted odds ratio (OR) for prevalent hepatic steatosis per 10‐kg lower HGS was 1.21 (95% CI: 1.04–1.40; p = 0.014) (Figure 2).

FIGURE 2.

Adjusted^# regression line (95% CI) showing the association of handgrip strength (HGS) with the odds ratio for the risk of prevalent hepatic steatosis (n = 2700). ^#Linear regression adjusted for age, sex, fat mass, fat‐free mass, self‐reported type 2 diabetes, hypertension, smoking status, alcohol consumption and use of lipid‐lowering medications.

3.4. Associations between HGS values and LFC stratified by BMI

The association of low HGS with LFC markedly differed among individuals with normal weight, overweight or obesity (Figure 3). Specifically, obese subjects with low HGS had a mean of 1.58% (95% CI: 0.18–2.98; p = 0.027) higher LFC, compared to obese with high HGS. In the overweight group, subjects with low HGS had a mean of 1.16% (95% CI: 0.474–1.848; p = 0.001) higher LFC compared to those with high HGS. On the other hand, there was no significant association between low HGS and LFC in the normal weight group. We did not find any significant association between low HGS and serum liver enzyme levels (ALT, AST, or GGT) in any of the three BMI groups.

FIGURE 3.

Adjusted^# mean (95% CI) of the association of high and low handgrip strength (HGS) defined by the fat‐free mass specific median* with liver fat content (LFC), stratified by increasing body mass index (normal weight, overweight or obesity). ^# Linear regression adjusted for age, sex, self‐reported type 2 diabetes, hypertension, smoking status, alcohol consumption, and use of lipid‐lowering medications. *The high and low fat‐free mass specific median of handgrip strength (HGS) were calculated as a HGS cutoff for each specific level of fat‐free mass.

Compared to participants with high HGS in the same BMI group, obese and overweight participants with low HGS had a higher chance of prevalent hepatic steatosis with an adjusted OR of 1.74 (95% CI: 1.15–2.62; p = 0.01) for obese subjects and 1.32 (95% CI: 1.01–1.74; p = 0.045) for overweight subjects, respectively. Conversely, the risk of prevalent hepatic steatosis was not significantly different between low HGS and high HGS in the normal weight group (adjusted OR = 1.20 [95% CI: 0.70–2.05; p = 0.51]).

4. DISCUSSION

In this large cross‐sectional population‐based study of adult German individuals, lower HGS (i.e., a marker of muscular fitness) were significantly associated with greater LFC (as measured by MRI‐PDFF), and higher serum GGT as well as AST levels, independent of age, sex, fat mass, fat‐free mass, self‐reported T2D, hypertension, smoking status, alcohol consumption, and use of lipid‐lowering medications. Individuals with low HGS were more likely also to have hepatic steatosis (defined as MRI‐PDFF ≥5.1%). Notably, after BMI stratification, we found that obese subjects with low HGS had a 74% higher chance of having hepatic steatosis when compared to obese subjects with high HGS. This association was also significant in overweight individuals with an odds ratio half as high than in obese individuals. In contrast, in individuals with normal body weight we did not observe any significant association between low HGS and risk of prevalent hepatic steatosis. Likewise, there were no significant associations between low HGS and serum liver enzyme concentrations in the aforementioned stratified analyses.

Our findings are in line with a previous study of our group (Zinterl et al., 2022) in which we observed similar associations for cardiorespiratory fitness. In that study, a lower cardiorespiratory fitness assessed by peak oxygen uptake was significantly associated with both greater LFC and higher serum GGT concentrations, as well as a higher chance for prevalent hepatic steatosis (Zinterl et al., 2022). Similarly to our current analyses, this relationship was more pronounced in overweight and obese participants (Zinterl et al., 2022). It is important to underline that while the results were similar, the assessment of HGS, instead of peak oxygen uptake, is easier and a more economic diagnostic tool (Roberts et al., 2011) for assessing physical fitness. The main findings of the present study are summarized in the central figure.

4.1. In the context of the published literature

In agreement with our findings, a cross‐sectional analysis of the Lanxi cohort (Gan et al., 2020) with 3536 Chinese participants, aged 18–80 years, showed that both lower weight‐adjusted HGS and lower skeletal muscle mass were independently associated with a higher risk of prevalent NAFLD as detected by ultrasound. The coexistence of low muscle mass and low muscle strength further increased the risk for prevalent NAFLD (Gan et al., 2020).

On the other hand, contrary to our findings, a cross‐sectional analysis of the UK Biobank study (Linge et al., 2021) with 5326 subjects aged 40–69 years, showed no association between HGS and presence of NAFLD, as detected by MRI‐PDFF. While our overall prevalence of NAFLD was 40%, it was 23% in the UK Biobank study, which suggests that the UK Biobank study might exhibit a healthy volunteer bias (Fry et al., 2017), which may have influenced the results. Moreover, the analysis of the UK Biobank study (Linge et al., 2021) (contrary to ours) used absolute HGS values that were not corrected for any body size or composition. The prevalence of overweight observed in subjects with NAFLD was almost two times greater than in those without NAFLD (Linge et al., 2021), which might have partly masked the relative loss of HGS in those subjects (Fogelholm et al., 2006; Lawman et al., 2016).

While there are several cross‐sectional studies (Han et al., 2022) examining the association between HGS and NAFLD, only two prospective studies exploring the association between HGS and NAFLD have been recently published. The first prospective study (Petermann‐Rocha et al., 2022), based on a large cohort (n = 333,295) of UK Biobank participants, aged 37–73 years at baseline and a median follow‐up of 10 years, showed that lower HGS was significantly associated with higher risk of developing severe NAFLD (defined as hospital admission or death due to NAFLD or nonalcoholic steatohepatitis) (Petermann‐Rocha et al., 2022). Further analysis also hypothesized that clinical interventions improving muscle mass might be a protective factor against the development of severe NAFLD (Petermann‐Rocha et al., 2022). Considering the strengths of that study with its longitudinal, prospective design with a large number of individuals, it is important to consider that the use of MRI in our analyses reported a significant association between lower HGS and the presence of hepatic steatosis already for the early stages of NAFLD in the general population. In line with our results, the second published prospective study (Xia et al., 2021) of Xia et al., based on the data from a general Chinese population (n = 14,154) with a mean follow‐up of 3.2 years, indicates that relative HGS was inversely associated with the risk of developing incident NAFLD (as detected by ultrasound) (Xia et al., 2021).

Finally, in line with our results, a cross‐sectional study (Lee et al., 2022) of 662 South Korean adults showed that the predicted probability of low HGS increased across higher levels of serum GGT. On the other hand, and contrary to our findings, another cross‐sectional analysis (Chung et al., 2020) with 1075 South Korean older men (≥50 years) and postmenopausal women with T2D showed that lower HGS was associated with lower serum ALT concentrations.

4.2. Potential mechanisms for the observed associations

Previous studies (Trenell, 2015) indicated that the effects of a sedentary lifestyle are a result of down and upregulated pathways, different to those linked with regular physical activity. Low muscle activity, following a more sedentary lifestyle, might result in changes related to the normal body physiology, precipitating overweight/obesity, and coexisting systemic insulin resistance, dysglycemia, and atherogenic dyslipidemia (Trenell, 2015), all of them promoting the development of hepatic steatosis. Otherwise, our statistical regression models considered the influences of fat mass, fat‐free mass, and plasma glucose levels, thereby suggesting a possible direct adverse effect of low levels of physical activity on risk of hepatic steatosis.

NAFLD and the decrease in muscular mass and strength shared similar pathophysiological mechanisms (Kim et al., 2019), which might explain the observed results of our study. The main potential mechanisms are physical inactivity, visceral obesity, skeletal muscle insulin resistance, low‐grade chronic inflammation and dysregulated secretion of hepatokines and myokines (Kim et al., 2019). Skeletal muscle insulin resistance, as a result from fat tissue infiltration in the muscle, leads to muscle degradation and mitochondrial dysfunction with increased oxidative stress (Kim et al., 2019). Additionally, the missing insulin effect on the skeletal muscle may lead to a decrease in protein synthesis, which results in further muscle loss and may aggravate systemic insulin resistance (Cleasby et al., 2016). Importantly, insulin resistance may lead to a reduced activity of the insulin signaling cascade, thereby reducing the insulin dependent glucose transporter‐4 (GLUT4) translocation and/or expression (Leguisamo et al., 2012). As the skeletal muscle is responsible for the main postprandial absorption of glucose, reduced GLUT4 expression would result in lower skeletal muscle glucose uptake, which would increase circulating levels of plasma glucose (Kim et al., 2019). This glucose overload would lead to increased hepatic de novo lipogenesis, thus resulting in increased LFC (Kim et al., 2019). Additionally, visceral obesity is associated with higher circulating free fatty acids, which accumulate in the skeletal muscle and further exacerbate systemic insulin resistance, and also inhibit the hepatic insulin signaling, thereby promoting hepatic insulin resistance (Kim et al., 2019). Therefore, both systemic and hepatic insulin resistance would be involved in the development of NAFLD. It is worth noting that, it has been demonstrated that hepatic steatosis is associated with insulin resistance in skeletal muscle rather than in the liver (Kato et al., 2015), thus highlighting the key role of the skeletal muscle in the development and progression of NAFLD (Kim et al., 2019). Finally, the dysregulated secretion of myokines, due to a loss of muscular fitness, can aggravates the aforementioned metabolic disorders, such as insulin resistance and higher circulating free fatty acids, and thus influence the development of hepatic steatosis (Kim et al., 2019; Yang et al., 2022).

Contrary to the liver specific ALT, AST exists in various tissues (Li et al., 2020). The higher serum AST levels observed in subjects with low HGS might be the result of coexisting damage to different tissues, including the skeletal muscle and the liver (Li et al., 2020). GGT is mainly located in the cell membrane and plays a role in the maintenance of intracellular reduced glutathione, which is a major antioxidant agent (Li et al., 2020). Consequently, our observed higher serum GGT concentrations in subjects with low HGS might be the result of higher oxidative stress in the liver that results from progression of hepatic steatosis to nonalcoholic steatohepatitis and fibrosis.

4.3. Study limitations

There are some important limitations of our study that need to be mentioned. Firstly, the cross‐sectional design, which does not permit conclusions of causality and temporality, so future longitudinal analyses and replication studies are needed. Secondly, our study sample included only Caucasian individuals; therefore, it may not be representative of other ethnic groups. Thirdly, we only had self‐reported information on whether individuals were taking part in leisure time exercise or not. A more objective and detailed survey of physical inactivity and sedentary time would have contributed to adjust more precisely for this potential confounding factor. Fourthly, liver biopsy, which is the gold standard method for diagnosing and staging hepatic steatosis, was not performed in our study, as it would not be ethically feasible in healthy individuals with fairly normal serum liver enzyme levels. Finally, although we incorporated several potential confounding factors in our multivariable adjusted regression models, we cannot exclude unmeasured or unknown residual confounders.

The major strengths of our study include the large number of individuals (n = 2700) with a wide age range (21–90 years) and the standardized measurement of HGS. Additionally, we used MRI‐PDFF for the measurement of LFC, which is now considered to be the most accurate imaging technique to non‐invasively diagnose and quantify hepatic steatosis (Piazzolla et al., 2020).

5. CONCLUSIONS

Our results from a large community‐based cohort showed a strong association of lower HGS with greater MRI‐measured LFC and higher risk for prevalent hepatic steatosis, particularly in individuals with overweight or obesity. Further studies are needed to better understand whether these findings may reflect the adverse effects of a sedentary lifestyle on the liver–“the sedentary's liver”.

AUTHOR CONTRIBUTIONS

Till Ittermann analyzed the data and is the guarantor of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Claudius Mayer wrote the manuscript. Marcello Ricardo Paulista Markus have contributed to conception and design of the manuscript. Till Ittermann, Sabine Schipf, Stefan Gross, Simon Kim, Jan Schielke, Robin Bülow, Jens‐Peter Kühn, Markus M. Lerch, Henry Völzke, Stephan Burkhard Felix, Martin Bahls, Giovanni Targher, Marcus Dörr and Marcello Ricardo Paulista Markus have contributed with substantial interpretation of the data and have critically revised the manuscript for important intellectual content. All authors gave final approval for the manuscript.

CONFLICT OF INTEREST STATEMENT

Giovanni Targher is supported in part by grants from the University School of Medicine of Verona, Verona, Italy. All other authors have no conflicts of interest to disclose related to this manuscript.

CONSENT FOR PUBLICATION

All study participants gave written informed consent to participate in this study, and having their results published as part of this study.

Supporting information

Supporting Information S1

DATA AVAILABILITY STATEMENT

The datasets generated during and/or analyzed during the current study are not publicly available due to data protection aspects but are available in an anonymized form from the corresponding author on reasonable request.