Myosteatosis: Diagnosis, pathophysiology and consequences in metabolic dysfunction-associated steatotic liver disease

Summary

Metabolic dysfunction-associated steatotic liver disease (MASLD) is associated with an increased risk of multisystemic complications, including muscle changes such as sarcopenia and myosteatosis that can reciprocally affect liver function. We conducted a systematic review to highlight innovative assessment tools, pathophysiological mechanisms and metabolic consequences related to myosteatosis in MASLD, based on original articles screened from PUBMED, EMBASE and COCHRANE databases. Forty-six original manuscripts (14 pre-clinical and 32 clinical studies) were included. Microscopy (8/14) and tissue lipid extraction (8/14) are the two main assessment techniques used to measure muscle lipid content in pre-clinical studies. In clinical studies, imaging is the most used assessment tool and included CT (14/32), MRI (12/32) and ultrasound (4/32). Assessed muscles varied across studies but mainly included paravertebral (4/14 in pre-clinical; 13/32 in clinical studies) and lower limb muscles (10/14 in preclinical; 13/32 in clinical studies). Myosteatosis is already highly prevalent in non-cirrhotic stages of MASLD and correlates with disease activity when using muscle density assessed by CT. Numerous pathophysiological mechanisms were found and included: high-fat and high-fructose diet, dysregulation in fatty acid transport and ketogenesis, endocrine disorders and impaired microRNA122 pathway signalling. In this review we also uncover several potential consequences of myosteatosis in MASLD, such as insulin resistance, MASLD progression from steatosis to metabolic steatohepatitis and loss of muscle strength. In conclusion, data on myosteatosis in MASLD are already available. Screening for myosteatosis could be highly relevant in the context of MASLD, considering its correlation with MASLD activity as well as its related consequences.

Key points.

• •This systematic review highlights the high prevalence of myosteatosis even in non-cirrhotic stages of MASLD.
• •Magnetic resonance imaging and spectroscopy are by far the most accurate assessment tools to characterise muscle phenotype.
• •Pathophysiological mechanisms included high-fat and high-fructose diet, dysregulation of fatty acid transport and ketogenesis, mitochondrial dysfunction, endocrine disorders and impaired microRNA122 pathway signalling.
• •Potential consequences related to myosteatosis are insulin resistance, MASLD progression from steatosis to steatohepatitis and loss of muscle strength.
• •Some pre-clinical and clinical data were conflicting and further large sample studies are required to investigate the muscle-liver-adipose tissue axis.

Introduction

Metabolic dysfunction-associated steatotic liver disease (MASLD), for which there is currently no approved treatment, is the most common liver disease worldwide. This entity includes hepatic steatosis, defined as a hepatic triglyceride content exceeding 5% of the total liver weight,¹ metabolic dysfunction-associated steatohepatitis (MASH), fibrosis and eventually cirrhosis and hepatocellular carcinoma² occurring in the context of metabolic comorbidities (overweight, metabolic syndrome, type 2 diabetes).³ MASLD prevalence is highly correlated to overweight, insulin resistance (IR) and abdominal obesity.⁴ Indeed, MASLD affects approximately 70% of overweight adults worldwide⁵ and up to 90% of those with morbid obesity.⁶ However, it currently remains difficult to identify individuals with MASLD at risk of progressing from isolated steatosis to MASH.

Liver steatosis results from several pathophysiological mechanisms including excessive dietary fatty acid intake,⁷ increased adipose tissue-related lipolysis,⁸ lipogenesis,⁹ altered mitochondrial fatty acid oxidation¹⁰ or even decreased hepatic lipid secretion.⁸

Skeletal muscle changes associated with MASLD correlate with all-cause mortality.¹¹ Those muscle changes include two main entities: sarcopenia and myosteatosis. Sarcopenia is defined by a loss of muscle mass and function while myosteatosis is defined by an excessive muscle lipid content.^12,13 Myosteatosis is frequent in cirrhosis and hepatic encephalopathy.¹⁴ While pathophysiological mechanisms linking hepatic encephalopathy and sarcopenia have been investigated deeply,¹⁵ the role of ammonia in the pathogenesis of myosteatosis is unknown. One hypothesis is that ammonia detoxification by glutamine synthase is decreased in fatty infiltrated skeletal muscles. However, decreasing muscle lipid content in a mouse model of MASH did not improve ammonia metabolism.¹⁶ Therefore, the relationship between ammonia and myosteatosis will not be further discussed in this review, considering the current lack of data.

Myosteatosis is also associated with MASLD at earlier disease stages.^12,[17], [18], [19] It has even been proposed as a key driver in MASLD pathogenesis.²⁰ However, myosteatosis is a highly unspecific muscle change seen in physiological (athletes, ageing) and pathological conditions (cachexia, neuromuscular degenerative diseases).[21], [22], [23], [24] Intramyocellular lipid content (IMLC) is also sex-dependent and is physiologically higher in females.²⁵ Muscle fat increases in athletes due to increased oxidative capacities.²¹ Indeed, chronic endurance training increases lipid storage in highly insulin-sensitive muscle cells as fuel for oxidative metabolism.²¹ This observation is referred to as the “athlete’s paradox” considering that myosteatosis is, in contrast, correlated with decreased oxidative capacities in insulin-resistant individuals with obesity.^21,26

Fat may accumulate outside (intermuscular fat) or inside muscle fascia (intramuscular fat). Intramuscular fat accumulates in the interstitium (extramyocellular) or in myocytes (intramyocellular)²⁷ (Fig. 1). It preferentially affects some muscles according to their oxidative capacities. IMLC is reported to be higher in oxidative muscles and to be a ready source of substrate for β-oxidation (βOX).²⁸ This observation is debated²⁹ as previous studies reported an increased IMLC in glycolytic muscles due to re-esterification of free fatty acids (FFAs) as a consequence of decreased mitochondrial oxidative phosphorylation.³⁰

Fig. 1.

Potential pathophysiological mechanisms, phenotypic description and consequences of myosteatosis in MASLD.

MASH, metabolic dysfunction-associated steatohepatitis; MASLD, metabolic dysfunction-associated steatotic liver disease.

Myosteatosis has a negative impact on muscle function, likely related to lipotoxicity, limited neuromuscular activation, impaired muscle blood flow and increased local inflammation.^22,31 Mechanically, fat infiltration reduces the pennation angle defined by the wrong alignment of muscle fibres and the muscle-force axis.³²

The pathophysiology of myosteatosis in MASLD as well as its related consequences remain unclear. Therefore, we performed a systematic review on myosteatosis in MASLD with a focus on diagnosis, pathophysiology and the consequences of its presence in MASLD.

Criteria and analysis for the systematic review

We combined PUBMED, EMBASE, COCHRANE databases from their inception up to December 31, 2022. The search terms were based on subject terms related to myosteatosis and MASLD. However, MASLD is a definition that emerged from a recent nomenclature consensus.³³ Search terms used its previous nomenclature, namely nonalcoholic fatty liver disease (NAFLD) and metabolic dysfunction-associated fatty liver disease (MAFLD). Those terms included “NAFLD and muscle fat”, “NAFLD or muscle fat”, “NAFLD and myosteatosis”, “NAFLD or myosteatosis”, “MAFLD and muscle fat”, “MAFLD or muscle fat”, “MAFLD and myosteatosis”, “MAFLD or myosteatosis”.

We did not restrict the search by region, language or period of time. We only included original manuscripts for this systematic review and excluded narrative and systematic reviews, as well as editorials and supplements.

The search of the computerised database turned up a total of 1,323 articles (478 on PUBMED, 779 on EMBASE and 66 on COCHRANE). Among these manuscripts, 518 were removed before screening as they were reviews, editorials or abstracts (Fig. 2). A total of 805 original manuscripts were retained and screened for inclusion. 694 of these original manuscripts were secondarily excluded as they were considered irrelevant or with the full online version unavailable. Original manuscripts were considered irrelevant if skeletal muscle fat content in individuals with MASLD was not specifically assessed or was assessed in non-skeletal muscles (e.g. myocardium). Hence, 111 entries were retained but included 65 duplicates. In total, 46 original manuscripts were included for this systematic review, including 14 pre-clinical and 32 clinical studies. The details of the inclusion process are summarised in the flowchart (Fig. 2). The details of included manuscripts are summarised in Table 1 for pre-clinical studies and Table 2 for clinical studies.

Fig. 2.

Flow chart of the selection of original manuscripts for inclusion in this systematic review.

MASLD, metabolic dysfunction-associated steatotic liver disease.

Table 1.

Original manuscripts on pre-clinical studies included in this systematic review listed in alphabetical order by first authors.

First author	Manuscript title	Journal	Year	Model	N diet/treatment	Investigated muscles	Investigation techniques of muscle lipid content
Brackmann C. et al.	Non-linear microscopy of lipid storage and fibrosis in muscle and liver tissues of mice fed high-fat diets	Journal of Biomedical Optics	2010	C57BL/6J mice	6 mice per group HFD vs. ND	Soleus diaphragm	Electron microscopy
Camporez J. et al.	Anti-inflammatory effects of oestrogen mediate the sexual dimorphic response to lipid-induced insulin resistance	Journal of Physiology	2019	C57BL/6J mice	8 mice per group ♂ vs. ♀ HFD vs. ND matched for age ♂ vs. ♀ HFD vs. ND matched for body weight ♂ vs. ♂ Estradiol vs. vehicle	Gastrocnemius soleus	1H-MRS tissue lipid extraction
Ceddia R. et al.	The PGE2 EP3 receptor regulates diet-induced adiposity in male mice	Endocrinology	2016	C57BL/6J mice EP3+/+vs. EP3-/-	7-10 mice per group EP3+/+ HFD vs. ND EP3-/- HFD vs. ND	Gastrocnemius soleus	Tissue lipid extraction
Chai C. et al.	Metabolic circuit involving free fatty acids, microRNA 122, and triglyceride synthesis in liver and muscle tissues	Gastroenterology	2017	C57BL/6J mice	3-14 mice per group CL316243 vs. control CL316243 ± antagomiR-122 antagomiR-122 vs. antagomiR-18	Not available	Optical microscopy
Hennige A. et al.	Enforced expression of protein kinase C in skeletal muscle causes physical inactivity, fatty liver and insulin resistance in the brain	Journal of Cellular and Molecular Medicine	2010	MLC-PKC-β2 transgenic C57BL/6 mice	4-8 mice per group 6 months old transgenic mice vs. WT on ND 3 months old transgenic mice HFD vs. ND	Tibialis muscles	MRI tissue lipid extraction
Imai N. et al.	Up-regulation of thioesterase superfamily member 2 in skeletal muscle promotes hepatic steatosis and insulin resistance in mice	Hepatology	2022	C57BL/6J S-Them2-/-	4 mice per group HFD vs. ND	Gastrocnemius	Optical microscopy tissue lipid extraction
Maj M. et al.	Consumption of high-fructose corn syrup compared with sucrose promotes adiposity and increased triglyceridemia but comparable NAFLD severity in Juvenile Iberian Pigs	Journal of Nutrition	2021	Iberian pigs	7 mice per group HS vs. HFCS	Longissimus dorsi	Optical microscopy
Marecki J. et al.	Hyperinsulinemia and ectopic fat deposition can develop in the face of hyperadiponectinemia in young obese rats	Journal of Nutritional Biochemistry	2011	Male Sprague-Dawley rats	10 mice per group TEN HFD vs. ND	Gastrocnemius	CT optical microscopy
Meneses M. et al.	Distinct impacts of fat and fructose on the liver, muscle, and adipose tissue metabolome: An integrated view	Frontiers in Endocrinology	2022	C57BL/6J	7 mice per group ND vs. HFD vs. HFrD	Gastrocnemius	1H-MRS tissue lipid extraction
Nachit M. et al.	Myosteatosis rather than sarcopenia associates with nonalcoholic steatohepatitis in non-alcoholic fatty liver disease preclinical models	Journal of Cachexia, Sarcopenia and Muscle	2020	foz/foz mice	48 foz/foz - 89 WT ND vs. HFD vs. HFFD	Erector spinae, quadratus lumborum, psoas	CT optical microscopy
Pichon C. et al.	Impact of L-ornithine L-aspartate on nonalcoholic steatohepatitis-associated hyperammonemia and muscle alterations	Frontiers in Nutrition	2022	foz/foz mice	6-7 mice per groupfoz/foz vs. WTND vs. HFDfoz/foz HFD +/- LOLA	Erector spinae/quadratus lumborumPsoas muscle	CT
						Quadriceps	Optical microscopy tissue lipid extraction
Preuss C. et al.	A new targeted lipidomics approach reveals lipid droplets in liver, muscle and heart as a repository for diacylglycerol and ceramide species in non-alcoholic fatty liver	Cells	2019	C57BL/6 SREBP-1c Tg mice	6 mice per groupC57B16 vs. SREBP-1c Tg ND	Not available	Optical microscopy tissue lipid extraction
Spooner H. et al.	High-fructose, high-fat diet alters muscle composition and fuel utilisation in a juvenile iberian pig model of non-alcoholic fatty liver disease	Nutrients	2021	Iberian pigs	6-8 pigs per groupND vs. ND + probiotics vs. HFFD vs. HFFD + probiotics	Longissimus dorsi	Tissue lipid extraction
Zhang W. et al.	Muscular G9a regulates muscle-liver-fat axis by musclin under overnutrition in female mice	Diabetes	2020	G9a-/- mice	3 mice per groupHFDMus33 vs. vehicle	Tibialis anterior	Optical microscopy

♂, male; ♀, female; antagomiR, microRNA antagonist; HFCS, high fructose corn syrup; HFD, high-fat diet; HFFD, high-fat high-fructose diet; HFrD, high fructose diet; ¹H-MRS, proton magnetic resonance spectroscopy; HS, high sucrose; LOLA, L-ornithine L-aspartate; MLC PKC β2, MLC protein kinase C beta-2; mus33, musclin 33; NAFLD, nonalcoholic fatty liver disease; ND, normal diet; PGE2 EP3, prostaglandin E2 E receptor 3; SREBP-1c Tg, sterol regulatory binding protein 1c triglycerides; TEN, total enteral nutrition.

Table 2.

Original manuscripts on clinical studies included in this systematic review listed in alphabetical order by first authors.

First author	Manuscript title	Journal	Year	N	Investigated muscles	Investigation techniques of muscle lipid content
Bhanji RA. et al.	Differing impact of Sarcopenia and Frailty in Nonalcoholic Steatohepatitis and Alcoholic Liver Disease	Liver Transplantation	2019	136 MASLD/129 ALD	Psoas, paraspinal and abdominal wall	CT
Chasapi A. et al.	Can obesity-induced inflammation in skeletal muscle and intramuscular adipose tissue accurately detect liver fibrosis ?	Journal of Musculoskeletal Neuronal Interactions	2018	50	Rectus abdominis	Optical microscopy
Cree Green M. et al.	Nonalcoholic fatty liver disease in obese adolescent females is associated with multi-tissue insulin resistance and visceral adiposity markers	Metabolism Open	2019	73	Soleus	1H-MRS 13C-MRS
Ding L. et al.	Myosteatosis in NAFLD patients correlates with plasma cathepsin D	Biomolecular Concepts	2021	45	Multifidus, erector spinae	MRI-PDFF
Han E. et al.	Muscle fat contents rather than muscle mass determines nonalcoholic steatohepatitis and liver fibrosis in patients with severe obesity	Obesity	2022	104	Psoas	CT
Hsieh Y. et al.	Muscle alterations are independently associated with significant fibrosis in patients with nonalcoholic fatty liver disease	Liver International	2020	521	Psoas, paraspinal and abdominal wall	CT
Hsieh Y. et al.	Myosteatosis, but not sarcopenia, predisposes NAFLD subjects to early steatohepatitis and fibrosis progression	Clinical Gastroenterology and Hepatology	2022	87 MASH/251 MASLD	Psoas, paraspinal and abdominal wall	CT
Jang S. et al.	Elevated serum alpha-fetoprotein level in asymptomatic individuals: clinical features, outcome and association with body fat deposition	Hepatology	2021	137	Psoas	CT
Jones H. et al.	Polycystic ovary syndrome with hyperandrogenism is characterised by an increased risk of hepatic steatosis compared to nonhyperandrogenic PCOS phenotypes and healthy controls, independent of obesity and insulin resistance	Journal of Clinical Endocrinology and Metabolism	2012	22 PCOS/22 controls	Soleus, tibialis anterior	1H-MRS
Jun D. et al.	Association between low thigh fat and non-alcoholic fatty liver disease	Journal of Gastroenterology and Hepatology	2008	408	Thigh muscles	CT
Kato K. et al.	Ectopic fat accumulation and distant organ-specific insulin resistance in Japanese people with nonalcoholic fatty liver disease	PLOS One	2014	69	Soleus	1H-MRS
Kitajima Y. et al.	Age-related fat deposition in multifidus muscle could be a marker for nonalcoholic fatty liver disease	Journal of Gastroenterology	2010	333	Multifidus muscle	CT
Kitajima Y. et al.	Severity of non-alcoholic steatohepatitis is associated with substitution of adipose tissue in skeletal muscle	Journal of Gastroenterology and Hepatology	2013	208	Multifidus muscle	CT
Linge J. et al.	Adverse muscle composition is linked to poor functional performance and metabolic comorbidities in NAFLD	Journal of Hepatology Reports	2021	9,545	Thigh	MRI PDFF
Machado V. et al.	Liver and muscle in morbid obesity: the interplay of fatty liver and insulin resistance	PLOS One	2012	51	Deltoid	Optical microscopy
Mey T. et al.	B-hydroxybutyrate is reduced in humans with obesity-related NAFLD and displays a dose-dependent effect on skeletal muscle mitochondrial respiration in vitro	American Journal of Physiology, Endocrinology and Metabolism	2020	1	Human primary myoblasts	Real-time respirometry
Nachit M., Kwanten W. et al.	Muscle fat content is strongly associated with NASH: A longitudinal study in patients with morbid obesity	Journal of Hepatology	2021	184	Psoas, dorsal and abdominal muscles	CT
Nachit M., Lanthier N. et al.	A dynamic association between myosteatosis and liver stiffness: Results from a prospective interventional study in obese patients	Journal of Hepatology Reports	2021	48	Psoas, dorsal and abdominal muscles	CT
Nakajima T. et al.	Age is a negative, and visceral fat accumulation is a positive, contributor to hepatic steatosis, regardless of the fibrosis progression in non-alcoholic fatty liver disease	Journal of Gastroenterology and Hepatology Research	2012	60 MASLD/26 controls	Multifidus muscle	CT
Oh S. et al.	Therapeutic effect of hybrid training of voluntary and electrical muscle contractions in middle-aged obese women with nonalcoholic fatty liver disease: A pilot trial	Therapeutics and Clinical Risk management	2015	15	Hamstrings, quadriceps	1H-MRS
Oh S. et al.	Whole-body vibration for patients with nonalcoholic fatty liver disease: a 6-month prospective study	Physiological reports	2019	25	Quadriceps	1H-MRS
Oshida N. et al.	Urinary Levels of Titin-N Fragment, a Skeletal Muscle Damage Marker, are Increased in Subjects with Nonalcoholic Fatty Liver Disease	Scientific Reports	2019	153 MASLD/100 controls	Rectus femoris	US 1H-MRS
Pasco J. et al.	Fatty Liver Index and Skeletal Muscle Density	Calcified Tissue International	2022	403	Radial, tibial	CT
Preuss C. et al.	A new targeted lipidomics approach reveals lipid droplets in liver, muscle and heart as a repository for diacylglycerol and ceramide species in nonalcoholic fatty liver	Cells	2019	1	Vastus lateralis	Optical microscopy
Shida T. et al.	Skeletal muscle mass to visceral fat area ratio is an important determinant affecting hepatic conditions of nonalcoholic fatty liver disease	Journal of Gastroenterology	2018	366	Bilateral quadriceps	1H-MRS
Shida T. et al.	Clinical and anthropometric characteristics of non-obese non-alcoholic fatty liver disease subjects in Japan	Hepatology Research	2020	404	Thigh	1H-MRS
Shigiyama F. et al.	Characteristics of Hepatic Insulin-Sensitive Nonalcoholic Fatty Liver Disease	Hepatology Communications	2017	26 MASLD/5 controls	Tibialis anterior	1H-MRS
Smajis S. et al.	Metabolic effects of a prolonged, very-high-dose dietary fructose challenge in healthy subjects	American Journal of Clinical Nutrition	2020	11 MASLD/10 controls	Not available	1H-MRS
Tarantino G. et al.	Circulating levels of sirtuin 4, a potential marker of oxidative metabolism, related to coronary artery disease in obese patients suffering from nafld, with normal or slightly increased liver enzymes	Oxidative Medicine and Cellular Longevity	2014	234	Left biceps brachii	US
Tarantino G. et al.	Interferon-alpha2 but not interferon-gamma serum levels are associated with intramuscular fat in obese patients with nonalcoholic fatty liver disease	Journal of Translational Medicine	2019	80 MASLD/38 controls	Left biceps brachii	US
Zhang W. et al.	Fat accumulation, Liver Fibrosis, and Metabolic Abnormalities in Chinese Patients with Moderate/Severe Versus Mild Hepatic Steatosis	Hepatology communications	2019	160	Dorsal muscles	CT
Zhang W. et al.	Metabolic abnormalities, liver and body fat in American versus Chinese patients with non-alcoholic fatty liver disease	Journal of Gastroenterology and Hepatology Open	2022	101 American/160 Chinese	Dorsal muscles	CT

ALD, alcohol-related liver disease; ¹³C-MRS, carbon 13 magnetic resonance spectroscopy; MASH, metabolic dysfunction-associated steatohepatitis; MASLD, metabolic dysfunction-associated steatotic liver disease; MRS, magnetic resonance spectroscopy; MRI-PDFF, magnetic resonance imaging proton density fat fraction; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; PCOS, polycystic ovary syndrome; SVR, skeletal muscle mass to visceral fat area ratio; US, ultrasound.

In pre-clinical studies, two main MASLD mouse models were used: transgenic mice on a C57BL/6J background (8/14) and foz/foz mice (2/14). Skeletal muscle lipid content was mainly assessed by microscopy (9/14), tissue lipid extraction (8/14), CT (3/14) and MRI (2/14). In clinical studies, muscle lipid content was assessed by CT (14/32), MRI (12/32), ultrasound (4/32) and microscopy (4/32). Muscle biopsy was only performed in three studies for histology (2/32) and tissue lipid extraction (1/32). Liver biopsy was performed in 12 studies. Paravertebral (4/14 in pre-clinical and 13/32 in clinical studies) and lower limb (10/14 in preclinical and 13/32 in clinical studies) muscles were the most frequently assessed.

Herein, we summarise the results of this systematic review in terms of the epidemiology of myosteatosis in MASLD, assessment techniques for myosteatosis, its pathophysiology and related consequences. We deliberately do not address the possible treatment of myosteatosis, which is beyond the scope of this work.

Myosteatosis prevalence in MASLD: Way more frequent than sarcopenia

The prevalence of myosteatosis according to MASLD severity remains poorly known considering the lack of histological assessment of liver disease. The prevalence of myosteatosis also remains a subject of debate due to the questionable techniques used to assess it. In real-life cohorts of individuals with MASLD with no liver histology available, myosteatosis is reported as highly prevalent (27.6%) compared to sarcopenia (1.6%). One study reported that myosteatosis and sarcopenia coexisted in 14% of cases.³⁴ Myosteatosis is also reported in pre-clinical and clinical studies with liver histology as an early muscle change linked with disease severity in non-cirrhotic stages.^19,35 However, myosteatosis is also more frequent compared to sarcopenia (83% vs. 30%) in individuals with cirrhotic MASH waiting for a liver transplantation.³⁶

How to assess myosteatosis?

Two types of techniques are used for muscle lipid content assessment: imaging and histology.

Imaging techniques

The diagnosis is routinely based on imaging. Three imaging techniques have been used to assess myosteatosis: CT, MRI and ultrasound (Table S1).

CT is the most used imaging modality in the field and highlights muscle fat accumulation through a lower mean muscle radiodensity expressed in Hounsfield units (HU).^18,37 While HU cut-off values are available from cancer cohorts, wherein cut-offs are used as mortality predictors,³⁸ validated cut-offs are not clearly established for individuals with chronic liver diseases or cirrhosis, due to an obvious lack of systematic data. A cut-off of 43.14 HU for psoas muscle radiodensity at the level of the fourth to fifth lumbar vertebra was evidenced as a good predictor of 12-month mortality in individuals with cirrhosis.³⁹ However, it was not the case for predicting post-liver transplant mortality in another study.⁴⁰ Decreases in skeletal muscle radiodensity to <41 HU in normal weight individuals (BMI up to 24.9 kg/m²) and to <33 HU in overweight individuals (BMI ≥25 kg/m²) have been proposed as criteria for myosteatosis and are associated with higher complication rates in those listed for liver transplant.⁴¹ Hence, CT only indirectly evaluates muscle fat content by its relative impact on muscle density and does not enable discrimination of its microscopic location (Table S1).

MRI enables the assessment of muscle phenotype through qualitative and quantitative sequences. Quantitative MRI sequences directly measure the relative fat content expressed in percentages by proton density fat fraction (PDFF) on T1-weighted images by suppressing the signal of water.^42,43 Furthermore, proton magnetic resonance spectroscopy (MRS) enables measurement of the IMLC and the extramyocellular lipid content (EMLC) based on their differences in resonant frequencies after excitation of hydrogen nuclei.^9,24 Proton MRS also enables the relative and absolute quantitation of muscle lipid content expressed in percentages and mass units (mmol/kg), respectively.⁴⁴ Hence, the main advantage of MRI compared to other imaging techniques (ultrasound or CT) is spectroscopy as well as the absence of radiation, even if the radiation level of contrast-free CT is very low.¹⁸ Therefore, proton MRS is the first-choice technique for myosteatosis assessment, considering that it non-invasively provides both quantitative measures and localisation information on muscle lipid content (Table S1).

Ultrasound is a non-radioactive imaging technique that allows for semi-quantitative measurement of muscle fat infiltration based on echo intensity variations. In case of myosteatosis, the echo intensity or brightness of skeletal muscle increases compared to normal muscles and is positively correlated with muscle lipid content.⁴⁵ However, despite a good correlation between echo intensity and MRI PDFF, ultrasound is examiner-dependent, only correlates with EMLC assessed by proton MRS, and provides inaccurate semi-quantitative measures of myosteatosis like CT^23,45,46 (Table S1).

Microscopic assessment of muscle lipid content

Although skeletal muscle biopsy is rarely performed in routine practice, innovative microscopy techniques have been used to assess skeletal muscle lipid content. The most used histological staining for muscle in optical microscopy is haematoxylin and eosin. Haematoxylin and eosin enables the measurement of muscle fibre diameter and adipocytes easily, while immunofluorescence enables differentiation between muscle fibre types.⁴⁷ Laser-based non-linear microscopy offers molecular-specific three-dimensional imaging of potentially large fresh samples. Two non-linear microscopy techniques have been used to assess lipid content and fibrosis in both liver and skeletal muscles in a mouse model of MASLD: second harmonic generation microscopy for the study of structural proteins such as collagen, and coherent anti-Strokes Raman scattering microscopy for the study of lipids.⁴⁸ The correlation between measures of muscle lipid content by these microscopic techniques and imaging has never been studied. As reported in Table 2, muscle lipid content was histologically assessed on rectus abdominis and deltoid in morbidly obese individuals using haematoxylin and eosin staining in just two clinical studies.^49,50 An increase in both EMLC and IMLC was reported in this population.^49,50

Potential biomarkers of muscle damage associated with myosteatosis in individuals with MASLD

Biomarkers of skeletal muscle damage such as creatine kinase assess the severity of muscle destruction or rhabdomyolysis.⁵¹ Serum levels of creatine kinase correlate with lean muscle mass and physiologically decrease with age.⁵² However, serum levels of creatine kinase are not specific to skeletal muscle and there is currently no data on the impact of myosteatosis on its serum level in MASLD. Two other biomarkers of muscle damage and body composition were found to be associated with muscle fat content in MASLD in two studies, though confirmation in larger series is required.^53,54

The first is the urinary level of titin-N fragment. Titin is a key structural protein of the sarcomere. The concentration of titin N-terminal fragment in urine is a specific biomarker of muscle damage. A positive correlation between myosteatosis (assessed by echo intensity of the rectus femoris) and urinary levels of titin-N fragment has been reported.⁵³ Interestingly, its urinary level was also correlated with forearm muscle strength.⁵³ The second one is serum alpha-foetoprotein (AFP) level. AFP is a commonly used tumour marker for diagnosis and follow-up of hepatocellular carcinoma and germline cell tumours. Surprisingly, serum AFP levels are related to body composition in healthy individuals.⁵⁴ Healthy individuals with elevated serum AFP levels have lower liver and muscle fat contents assessed by CT compared to age- and sex-matched controls with normal serum AFP levels.⁵⁴ Mean psoas density remained an independent factor for elevated serum AFP levels even after adjustment for liver steatosis. However, there is no correlation between muscle density and serum AFP level.⁵⁴

Pathophysiology of myosteatosis in MASLD

Excessive lipid storage in skeletal muscle is mediated by several pathophysiological mechanisms in MASLD and can be characterised by an imbalance between three main physiological pathways: lipid intake, oxidation and export. There is little published data on this topic and available data is almost exclusively in end-stage liver disease.⁵⁵ In the following section we summarise potential pathophysiological mechanisms (Fig. 3).

Fig. 3.

Pathophysiological mechanisms involved in myosteatosis pathogenesis in MASLD.

MASLD, metabolic dysfunction-associated steatotic liver disease.

Fat- and fructose-enriched diets increase liver and muscle lipid contents

In the same way that diet causes steatosis, consumption of a high-fructose (HFrD) or a high-fat diet (HFD) is associated with myosteatosis.[56], [57], [58], [59], [60], [61], [62], [63] According to tracer experiments, the most important source of lipids in the steatotic liver is the insulin-resistant adipose tissue.⁶⁴ It is therefore logical to assume that FFAs from the adipose tissue also contribute to myosteatosis.

Fructose has been incriminated in the growing pandemic of MASLD worldwide considering its frequent use in the food industry, notably in the form of high-fructose corn syrup, and its well-known toxicity.⁶³ Indeed, during fructolysis, fructose is converted into glyceraldehyde-3-phosphate, a substrate for de novo hepatic lipogenesis, eventually leading to increased circulating FFAs. Furthermore, a HFrD induces a drop in the ATP/AMP ratio as increased cellular respiration flux increases the consumption of inorganic phosphate, subsequently increasing the risk of MASLD development and progression.⁵⁶ Fructose also exerts direct mitochondrial toxicity by generating reactive oxygen species-damaging mitochondrial DNA and impairing mitochondrial biogenesis.⁵⁷ This oxidative stress-mediated mitochondrial toxicity was reported in hepatocytes but also in myocytes.^57,58 Finally, decreased lipolysis in the liver secondary to a HFrD was also highlighted.⁵⁹ Hence, a HFrD increases liver lipid content not only via de novo lipogenesis but also via excessive lipid storage (by reducing lipolysis, mitochondrial content and fatty acid βOX in both the liver and skeletal muscle).

A HFrD induced more liver inflammatory degeneration and hence further promoted MASH compared to a HFD which induced more liver steatosis.⁶⁰ This pro-inflammatory role of fructose was reinforced by a reported association with higher muscle levels of interleukin (IL)1α.⁶¹ Both a HFrD and a HFD promote IR in the liver and skeletal muscle by impairing the insulin signalling pathway. However, previous data tend to demonstrate a deeper impact of a HFD on skeletal muscle insulin sensitivity, promoting the accumulation of other energetic substrates such as lipids.^60,61

Furthermore, the impact of a HFD on muscle lipid content but also on skeletal muscle insulin sensitivity is reported as highly dependent on the subtype of dietary lipid intake. Indeed, a high intake of saturated fat worsens IR while long-chain n-3 fatty acids enhance insulin sensitivity by promoting muscle mitochondrial oxidative capacities.⁶²

Dysregulation in intramyocellular fatty acid transport

The regulation of intramyocellular fat content is highly similar to that of muscle glucose content. Indeed, IMLC is increased by insulin though the PI3K-PKB-Akt signalling pathway and by muscle contraction though increasing cytosolic levels of AMP-activated protein kinase (AMPK).⁶⁵ It results in increased translocation of fatty acid transporters such as fatty acid translocase/CD36 and fatty acid transport protein.⁶⁵ Previous studies reported increased expression of CD36 and fatty acid transport protein in skeletal muscle of individuals with obesity and type 2 diabetes, arguing in favour of excessive fatty acid uptake promoting myosteatosis in addition to excessive dietary intake.⁶⁶

Dysregulation of myocyte mitochondria

When switching to a HFD, skeletal muscle rapidly increases lipid uptake to reduce serum levels of FFAs.⁶⁵ As a response, mitochondrial βOX increases preferably in oxidative muscle.³⁰ However, mitochondrial capacity is limited, especially in glycolytic muscles due to lower mitochondrial oxidative phosphorylation compared to oxidative fibres.³⁰ Glycolytic muscles are therefore at higher risk of myosteatosis compared to oxidative muscles.³⁰ Unfortunately, when excessive IMLC is chronic, the oxidative capacity of myocyte-related mitochondria decreases, eventually amplifying myosteatosis.⁶⁷ Expression of key regulators of lipid oxidation (peroxisome proliferator-activated receptor γ coactivator 1α and carnitine palmitoyltransferase 1) are also decreased in juvenile Iberian pigs fed a high-fructose, high-fat diet, resulting in increased muscle lipid content.⁶¹ Conversely, promoting βOX in skeletal muscle of mice decreases the impact of HFD on muscle lipid content. Indeed, downregulating the acyl-CoA thioestherase 2, a mitochondrial enzyme that decreases the availability of βOX substrates by decreasing the hydrolysis of long-chain fatty acyl-CoA in skeletal muscle, protects against HFD-induced myosteatosis.⁶⁸

However, the pathophysiological role of mitochondrial dysfunction in myosteatosis remains uncertain due to contradictory data. Indeed, recent studies do not associate IMLC with mitochondrial oxidative capacity. The mitochondrial function of skeletal muscle in individuals with MASLD was indirectly studied and compared to that of healthy individuals by phosphorus MRS assessing mitochondrial ATP production.^69,70 In parallel, IMLC in lower limb muscles was assessed by proton MRS.^69,70 Individuals with MASLD showed a higher level of liver steatosis but no difference in terms of mitochondrial function (as assessed by ATP measured by phosphorus MRS) and IMLC (evaluated by proton MRS).^69,70

Impaired ketogenesis

Impaired ketogenesis has been reported to be associated with increased hepatic steatosis in mice,⁷¹ while inducing ketogenesis reduces its content in humans.⁷² Impaired ketogenesis might then play a role in MASLD pathophysiology but also in IR. However, owing to conflicting data, it remains unclear whether hepatic ketogenesis in MASLD is impaired or not.⁷³ Interestingly, the administration of β-hydroxybutyrate in vitro induces a dose-dependent decrease in mitochondrial respiration of human myocytes, as assessed by real-time respirometry.⁷³ This observation highlighted a potential link between hepatic ketogenesis and peripheral insulin sensitivity through impaired skeletal muscle mitochondrial respiration.⁷³ However, this relationship has only been reported in one study and therefore further investigations focusing on this topic are required.

Endocrine disorders

Oestrogen deficiency and hyperandrogenism are endocrine disorders that promote the onset of myosteatosis in MASLD.

Oestrogen deficiency

Oestrogen is a sexual hormone involved in glucose and lipid metabolism but also in insulin sensitivity. Indeed post-menopausal women frequently face an increase in global fat mass exposing them to a higher risk of MASLD and related complications such as myosteatosis.⁷⁴ Conversely, oestrogen supplementation has been reported to significantly reduce liver and muscle fat in a male mouse model, leading to enhanced insulin sensitivity by decreasing protein kinase C theta activation.⁷⁵ Furthermore, oestrogen supplementation has anti-inflammatory properties by decreasing IL6, IL1β and tumour necrosis factor-α (TNFα) levels in plasma and white adipose tissue.⁷⁵

Hyperandrogenism

MASLD prevalence is increased in polycystic ovary syndrome, which is often associated with increased visceral fat content, IR and other features of metabolic syndrome.⁷⁶ IR in this syndrome is associated with hyperandrogenism itself. Indeed, liver fat content is increased in women with hyperandrogenic polycystic ovary syndrome compared to normoandrogenic women.⁷⁷ Hyperandrogenism is also positively correlated with IMLC.⁷⁷ This association between hyperandrogenism and MASLD has also been observed in a gender comparative study. Male gender is an independent risk factor for MASLD, potentially due to increased testosterone levels.⁷⁸ Moreover, increased serum testosterone levels, even within a physiological range, have also been reported to be a risk factor for MASH development in type 2 diabetes.⁷⁹ Hence, androgens might increase muscle lipid content by promoting MASH development though pro-inflammatory properties, but clinical data are contradictory and further investigations are required.^80,81

Impaired microRNA-122 pathway

MicroRNA-122 (miR-122) is a non-coding RNA expressed in the liver and potentially secreted into the bloodstream with local and systemic effects.⁸² Indeed, miR-122 is involved in lipid metabolism and several pathological conditions such as hepatitis C.⁸³ MiR-122 expression and active secretion by hepatocytes are induced in response to increased free and hepatic fatty acids via activation of the retinoic acid-related orphan receptor-α. MiR-122 decreases FFA synthesis and lipid storage by suppressing triglyceride synthesis in the liver and muscles.⁸⁴ MiR-122 also increases white adipose tissue lipolysis and βOX.⁸⁴ The expression and secretion of this physiological brake for triglyceride synthesis and storage are decreased in MASH.⁸⁵ However, the exact role of miR-122 as a regulator of lipid metabolism remains debated due to contradictory data possibly secondary to its circadian expression profile.^86,87 To our knowledge, the direct role of miR-122 in the pathophysiology of myosteatosis in MASLD has not been explored.

Three possible consequences of myosteatosis

Myosteatosis in MASLD has three potential metabolic and functional consequences that are reviewed below: IR, MASH and impaired skeletal muscle function (Fig. 1).

Myosteatosis and IR in humans: Is EMLC or IMLC the culprit?

The impact of muscle fat on insulin sensitivity in humans is a burning question that is currently difficult to answer considering the very limited data available and the lack of consistency across studies in terms of assessment techniques. Concerning intramuscular fat, total intramuscular fat assessed by CT positively correlates with IR in individuals with obesity and diabetes.²⁶ Intramuscular fat also mechanically impairs insulin sensitivity by reducing muscle blood flow and so insulin bioavailability.²⁶ However, this association is not reported in clinical studies assessing liver histology, IR and muscle density (Table 3).

Table 3.

Associations between liver phenotype (assessed histologically), muscle lipid content and insulin resistance (assessed by the homeostasis model assessment of insulin resistance) in clinical studies.

Associations investigated (myosteatosis vs. liver histology)	Reported associations	References	Assessment technique of myosteatosis	Association between myosteatosis and IR
Myosteatosis - liver steatosis	Muscle lipid content ↑ when liver lipid content ↑	Han E. Obesity. 2022	CT (SMFI)	–
		Kitajima Y. J Gastroenterol. 2010	CT (IMAC)	–
		Nachit M., Kwanten W. et al. J Hepatol. 2021	CT (SMFI)	–
	Muscle lipid content ↔ when liver lipid content ↑	Hsieh Y. Clin Gastroenterol Hepatol. 2022	CT (MA)	No association
		Machado M. PLOS One. 2012	OM	No association with IMLC & EMLC
	Muscle lipid content ↓ when liver lipid content ↑	Hsieh Y. Liver Int. 2021	CT (MA)	No association
Myosteatosis – total MASLD activity	Muscle lipid content ↑ when MASLD activity ↑	Han E. Obesity. 2022	CT (SMFI)	–
		Hsieh Y. Liver Int. 2021	CT (MA)	No association
		Hsieh Y. Clin Gastroenterol Hepatol. 2022	CT (MA)	No association
		Kitajima Y. J Gastroenterol Hepatol. 2013	CT (IMAC)	–
		Machado M. PLOS One. 2012	OM	No association with IMLC & EMLC
		Nachit M., Kwanten W. et al. J Hepatol. 2021	CT (SMFI)	–
Myosteatosis - lobular inflammation	Muscle lipid content ↑ when lobular inflammation ↑	Hsieh Y. Liver Int. 2021	CT (MA)	No association
		Hsieh Y. Clin Gastroenterol Hepatol. 2022	CT (MA)	No association
		Machado M. PLOS One. 2012	OM	No association with IMLC & EMLC
		Nachit M., Kwanten W. et al. J Hepatol. 2021	CT (SMFI)	–
Myosteatosis - ballooning	Muscle lipid content ↑ when hepatocyte ballooning ↑	Han E. Obesity. 2022	CT (SMFI)	–
		Hsieh Y. Liver Int. 2021	CT (MA)	No association
		Hsieh Y. Clin Gastroenterol Hepatol. 2022	CT (MA)	No association
		Nachit M., Kwanten W. et al. J Hepatol. 2021	CT (SMFI)	–
Myosteatosis - fibrosis	Muscle lipid content ↑ when liver fibrosis ↑	Han E. Obesity. 2022	CT (SMFI)	–
		Hsieh Y. Liver Int. 2021	CT (MA)	No association
		Hsieh Y. Clin Gastroenterol Hepatol. 2022	CT (MA)	No association
		Machado M. PLOS One. 2012	OM	No association with IMLC & EMLC
		Nachit M., Kwanten W. et al. J Hepatol. 2021	CT (SMFI)	–

EMLC, extramyocellular lipid content; IMAC, intramuscular adipose tissue content (density of the region of interest of the muscle/density of the region of interest of subcutaneous fat); IMLC, intramyocellular lipid content; MA, muscle attenuation; MASLD, metabolic dysfunction-associated steatotic liver disease; SMFI, skeletal muscle fat index.

When measured separately by MRS, IMLC and EMLC do not seem to have the same impact on insulin sensitivity.⁸⁸ Indeed, an association between IMLC assessed by proton MRS and IR is reported in one cohort of lean individuals with type 2 diabetes matched for age, sex and BMI.⁸⁹ On the flip side, a drop in IMLC induced by physical exercise has been reported to be correlated with improved insulin sensitivity in individuals with MASLD independently of body weight and decreased liver steatosis.^90,91 However, this association has not been reported in other diabetic cohorts focusing on the topic or in healthy athletes with increased IMLC.^21,70,92 On the contrary, in another study no association between EMLC, total adiposity and IR was observed.⁸⁹ Therefore, reducing the link between myosteatosis and IR to this lipotoxicity alone appears simplistic.

Intermuscular fat also impacts insulin sensitivity.⁹³ Indeed, an association between intermuscular fat invasively collected from the vastus lateralis of both diabetic and non-diabetic individuals with obesity and IR has been reported.⁹⁴ Interestingly, in the same study, exposing cell-cultured myotubes to this intermuscular fat from non-diabetic individuals with obesity increases the IMLC, specifically by increasing intramyocellular diacylglycerol (DAG) levels.⁹⁴ Hence, intramuscular fat is therein reported as a consequence of intermuscular fat exposure.⁹⁴ One hypothesis linking inter and intramuscular fat is that initially a decrease in the insulin sensitivity of intermuscular fat leads to lipolysis in intermuscular adipose tissue.⁹⁴ Therefore, intermuscular fat releases FFAs into the microenvironment of myocytes promoting EMLC and secondarily IMLC expansion.⁹⁴

Muscle density correlates with MASH

Concerning the correlation between liver histology and muscle fat, previous data reported a positive correlation between muscle lipid content, MASLD activity (ballooning and lobular inflammation)^19,50,[95], [96], [97], [98] and liver fibrosis^19,50,[96], [97], [98] (Table 3). Interestingly, the only study with liver and muscle histology reported a positive correlation exclusively between lobular inflammation and IMLC, ballooning not being assessed.⁵⁰ EMLC did not correlate with liver histology.⁵⁰ Data on the correlation between liver and muscle lipid contents are more conflicting^19,50,[96], [97], [98], [99] (Table 3). Interestingly, no correlation between BMI and muscle fat assessed by CT was reported in those studies.

Conversely, a decrease in muscle fat after weight loss induced by dietary changes or bariatric surgery and by physical exercise is associated with histological regression of liver steatosis and ballooning.^19,95 However, in all these studies comparing muscle and liver phenotypes, muscle lipid content was exclusively semi-quantitatively assessed by CT. Indeed, no study with liver histology assessed muscle lipid content by proton MRS in individuals with MASLD (Table 3). Therefore, the association between histological features of MASLD and each intramuscular lipid compartment (IMLC and EMLC) is a question currently unanswered. Furthermore, muscle density was often normalised by different parameters across studies (height, subcutaneous adipose tissue density) making an inter-study comparison hazardous (Table 3).

Myosteatosis might impact skeletal muscle function

Muscle function decay, defined by a decrease in muscle strength or performance, has been deeply investigated in sarcopenia with well-defined assessment tools.¹⁰⁰ Muscle function has been studied to a much lesser extent in myosteatosis with no assessment tools currently validated.

One retrospective study assessed the impact of myosteatosis on muscle function in a real-life cohort of individuals with MASLD and reported a similar decrease in muscle strength for both myosteatosis and sarcopenia.³⁴ Mixed muscle changes combining a high muscle lipid content and a low muscle volume have a synergic effect on muscle function with a greater loss of muscle strength.³⁴ However, volume and lipid content of anterior thigh muscles were assessed by MRI-PDFF.³⁴ Hence, there is no available data on the specific impact of IMLC and EMLC on muscle function. Furthermore, muscle function was assessed by questionnaires on walking pace and stair climbing habits which are poor assessment methods.

Other studies assessed the impact of muscle fat on muscle function. However, muscle function was either retrospectively poorly assessed by a deficit index with no functional impact of muscle fat,³⁶ or accurately assessed but not on the investigated muscles (muscle function measured on forearm muscles and lipid content by proton MRS on rectus femoris).⁵³ Further investigations accurately investigating the correlation between muscle lipid content, preferably assessed by proton MRS, and muscle function are required.

What promotes myosteatosis-related consequences in MASLD?

Two pathophysiological mechanisms linking myosteatosis to its potential complications cited above have been highlighted with this systematic review: lipotoxicity and disturbances in the muscle-liver-adipose tissue axis notably mediated by inflammation (Fig. 1).

Cytotoxicity related to intracellular lipids: The lipotoxicity concept

Many lipid species in MASLD cause hepatocyte toxicity and are involved in MASLD progression, such as saturated fatty acids and ceramide bodies.¹⁰¹ In a mouse model of MASLD, DAG accounted for the majority of the lipids that accumulated as lipid droplets in the myocyte cytosol.¹⁰² Indeed, cytosolic droplets of DAG have been reported to increase by fivefold in diet-induced MASLD.¹⁰² Cytosolic droplets of ceramide bodies also increase in these mice but in smaller proportions.¹⁰² However, at present, the relative contributions of these lipids to lipotoxicity in the muscle is not known. Lipotoxicity has two main consequences: IR and inflammation.

Skeletal muscle is a key organ in glucose homeostasis, removing up to 75% of all blood glucose content and is therefore extremely insulin sensitive.¹⁰³ FFAs can modulate muscle insulin sensitivity via several mechanisms.

Skeletal muscle mitochondrial βOX is decreased in overweight individuals with diabetes.^104,105 Intramyocellular lipids are oxidised into acetyl-CoA to produce ATP via the tricarboxylic acid cycle and oxidative phosphorylation. In case of acetyl-CoA accumulation secondary to increased βOX, glycolysis and pyruvic dehydrogenase are inhibited by negative feedback.¹⁰⁶ This leads to allosteric inhibition of hexokinase and eventually reduced myocyte glucose uptake, lower respiration rates and IR¹⁰⁷ (Fig. 4). Interestingly, restoring muscle βOX by administering an NAD+ precursor in overweight individuals with diabetes significantly improves insulin sensitivity.¹⁰⁵

Fig. 4.

Signalling pathways involved in the cellular lipotoxicity related to myosteatosis in MASLD.

AMPK, AMP-activated protein kinase; CPT1, carnitine palmitoyltransferase 1; DAG, diacylglycerol; GLUT4, glucose transporter 4; IL, interleukin; IRS, insulin receptor substrate; MAPK, mitogen-activated protein kinase; MASLD, metabolic dysfunction-associated steatotic liver disease; Myd88, myeloid differentiation factor 88; NLRP3, NOD-like receptor family, pyrin domain containing 3; PDK1, 3-phosphoinositide-dependent protein kinase 1; PGC1α, peroxisome proliferator-activated receptor gamma 1 α; PLIN5, perilipin 5; PI3K, phosphatidylinositol-3-kinase; PKC, protein kinase C; TLR4, Toll-like receptor 4; TNFα, tumour necrosis factor α.

Furthermore, some lipid molecules directly promote IR by interfering with glucose uptake.^108,109 Indeed, DAG and ceramide bodies, both products of triacylglycerol oxidation, halt the translocation of glucose transporter 4 (GLUT4) into the myocyte membrane.^108,109 In addition, DAG decreases the insulin-mediated activation of insulin receptor substrate-1 (IRS-1) by up to 30% by activating the protein kinase C theta.¹¹⁰ This protein kinase increases IRS-1 Ser307 phosphorylation which decreases IRS-1 tyrosine phosphorylation and IRS-1-associated PI3-kinase activation. This eventually decreases GLUT4 membrane expression¹¹⁰ (Fig. 4).

However, abnormal structural proteins of lipid droplets are also involved in this lipotoxicity. Perilipin 5 is a lipid droplet-associated protein expressed at the surface of cytosolic lipid droplets that regulates lipid hydrolysis and thus intracellular DAG and ceramide body levels.¹¹¹ It is also involved in the delivery of fatty acids to myocyte mitochondria to promote βOX and hence prevent myosteatosis.¹¹¹ Indeed, perilipin 5 knockout mice develop IR secondary to accumulation of ceramide bodies and triacylglycerol¹¹² (Fig. 4).

Impaired AMPK signalling also plays a key role in MASLD via several mechanisms.¹¹³ AMPK is a ubiquitous energy sensor involved in many homeostatic pathways depending on the tissue studied. It is composed of three subunits (α, β, γ) and is activated in case of low energy stores. In MASLD, hepatic AMPK is downregulated. Promoting AMPK signalling pharmacologically results in reduced serum inflammation by stimulating the anti-inflammatory function of macrophages.^114,115 This anti-inflammatory property is mediated by suppressing the expression of key pro-inflammatory mediators such as NF-κB,^116,117 IL6 or even TNFα.¹¹⁸ Furthermore, experimental studies on mouse models demonstrated that upregulation of AMPK signalling fights against IR specifically in skeletal muscle cells by promoting glucose uptake via increased GLUT4 translocation^119,120 (Fig. 4).

Eventually, FFAs have direct pro-inflammatory properties by activating innate immunity via the myeloid differentiation factor 2/Toll-like receptor 4 (TLR4) pathway.¹²¹ Myeloid differentiation factor 2/TLR4 not only recognises lipopolysaccharides present on the bacterial wall but also metabolism-associated molecular patterns, leading to the recruitment of intracellular adaptor protein myeloid differentiation factor 88. Myeloid differentiation factor 88 increases pro-inflammatory cytokine expression and release via the mitogen-activated protein kinase and NK-κB pathways.¹²² The increase in the intracellular level of NF-κB activates the NLRP3 (NOD-like receptor family, pyrin domain containing 3) inflammasome, promoting extracellular IL1β and IL18 release. Interestingly, NLRP3 participates in hepatic inflammation and hence MASLD pathogenesis and progression in mouse models¹²³ (Fig. 4).

The muscle-liver-adipose tissue axis is disrupted in individuals with MASLD and myosteatosis

The muscle-liver-adipose tissue axis is an inter-organ crosstalk involved in the metabolic regulation of all three organs involved.¹²⁴ Circulating mediators specifically involved in this crosstalk might originate from each tissue: hepatokines from the liver,^125,126 myokines from the skeletal muscle¹²⁷ and adipokines from the adipose tissue,¹²⁸ with autocrine, paracrine and endocrine properties. The expression and function of these mediators are disturbed in MASLD, promoting metabolic dysfunction including myosteatosis, eventually leading to liver and muscle function decay.[128], [129], [130] However, even if those mediators have been highly studied in MASLD these last years, there is only limited data available concerning the impact of myosteatosis on those mediators.

In the following section, we present the tissue-specific mediators whose expression levels are reported as specifically disturbed in individuals with MASLD and myosteatosis (Fig. 5).

Fig. 5.

Mediators of the muscle-liver-adipose tissue axis promoting myosteatosis in MASLD.

Adiponectin serum level decreases in MASLD, increasing liver and muscle lipid contents. Cathepsin D serum level increases in MASLD, promoting systemic inflammation. Musclin serum level decreases in MASLD, increasing liver lipid content. Seleprotein P serum level increases in MASLD, increasing liver and muscle lipid contents and decreasing adiponectin expression. Systemic inflammation and insulin resistance occurring in MASLD reciprocally promote an increase in muscle and liver lipid contents as well as adipose tissue expansion. IFNα2, interferon-α2; IL15, interleukin 15; IL1α, interleukin-1α; MASLD, metabolic dysfunction-associated steatotic liver disease; SeP, selenoprotein P; TGFβ, transforming growth factor β; TNFα, tumour necrosis factor-α.

Myosteatosis in individuals with MASLD impacts tissue-specific mediators of the muscle-liver-adipose tissue axis

Hepatokines

Hepatokines are proteins secreted into the circulation by hepatocytes with autocrine, paracrine and endocrine functions.¹²⁵ These mediators can interact with skeletal muscle and positively or negatively modify its metabolic functions.¹²⁶ Hepatokine secretion profiles are altered in MASLD, promoting inflammation, IR and potentially myosteatosis¹²⁹ (Fig. 5).

Selenoprotein P (SeP) is a glycoprotein mainly involved in the transport of selenium from the liver to peripheral tissues.¹³¹ SeP is increased in MASLD and promotes IR by altering insulin signalling.¹³² SeP increases phosphorylation of IRS-1 at Ser307 which, as mentioned earlier, downregulates tyrosine phosphorylation of IRS-1 and alters distal insulin signal transduction.¹³³ SeP also inhibits the AMPK/acetyl-CoA carboxylase pathway.¹³³ This inhibition results in increased intracellular lipid accumulation in hepatocytes and myocytes.¹³³ Furthermore, SeP might also promote IR by decreasing the expression of adiponectin, considering a negative correlation between serum levels of SeP and adiponectin reported in type 2 diabetes (Fig. 5).¹³⁴

Cathepsin D is a lysosomal enzyme mainly produced by hepatocytes. Cathepsin D is released from hepatocytes in case of liver steatosis as a consequence of lysosomal dysfunction, resulting in an increased serum level in MASLD.¹³⁵ It activates the TLR4 pathway, thereby increasing systemic and liver inflammation by upregulating the expression of pro-inflammatory cytokines (TNFα, CCL2).[136], [137], [138] Hence, cathepsin D might be involved in the progression of MASLD to MASH.¹³⁹ Furthermore, the serum levels of cathepsin D have been reported to be positively correlated with IR and myosteatosis (Fig. 5), but further investigations are required to determine its precise role in the pathogenesis of myosteatosis in MASLD.¹⁴⁰

Musclin (myokine)

Myokines are proteins secreted by the skeletal muscle with potential metabolic effects mediated by endocrine, paracrine or autocrine features. Serum levels of myokines vary according to the degree of physical activity.¹²⁷ The profile of secreted myokines is also disrupted in MASLD and has been reported to negatively modulate the liver-muscle axis. Musclin is a myokine that has been reported to be decreased in MASLD, with a direct impact on muscle lipid content (Fig. 5). Indeed, downregulating musclin expression by G9a, a histone methyltransferase that recruits a repressive marker (H3K9me2), increases muscle fat content.^130,141 On the contrary, upregulating muscle expression by suppressing G9a decreases muscle and liver fat contents via the FOXO1 pathway.¹³⁰ However, the pathways mediating its positive impact on liver and skeletal muscle remain unknown.

Adiponectin (adipokine)

Adipokines are mediators produced by adipocytes and have previously been described as major players in the development of IR in MASLD. These mediators are reported to be involved in several metabolic pathways through a highly regulated dynamic balance that is disturbed by adipose tissue expansion in MASLD¹²⁸ (Fig. 5). Adiponectin acts as a key regulator of energy metabolism in the liver and skeletal muscle by activating AMPK and PPAR (peroxisome proliferator-activated receptor)-α signalling pathways and thereby increasing fatty acid β- and ω-oxidation.¹⁴² This leads to enhanced insulin sensitivity by reducing liver and muscle lipid content.¹⁴³ Serum levels of adiponectin are significantly decreased in individuals with obesity and are negatively correlated with IR and type 2 diabetes.^144,145 However, this decrease in adiponectin serum levels is not reported as a key mechanism sufficient for IR development in MASLD.¹⁴⁶

Inflammation increases with myosteatosis and disrupts the muscle-liver-adipose tissue axis

The link between inflammation and myosteatosis in MASLD has been reported by several studies highlighting increased serum and tissue levels of inflammatory biomarkers in obese and non-obese individuals with increased muscle lipid content (Fig. 5). Indeed, serum levels of interferon-α2 (IFNα2), a cytokine involved in the early immune responses to viral infections, are increased in individuals with obesity and myosteatosis compared to healthy individuals.¹⁴⁷ Other members of the interferon family are not overexpressed. Furthermore, IFNα2 serum levels are inversely correlated with muscle lipid content assessed by ultrasound in these individuals and positively correlated with visceral adipose tissue.¹⁴⁷ However, the pathophysiological mechanism linking IFNα2 and muscle fat requires further investigation. Beside IFNα2 in individuals with obesity, serum levels of other inflammatory mediators are increased in overweight individuals with MASLD and myosteatosis, such as TNFα, IL1α and TGFβ.^61,[148], [149], [150]

Furthermore, histologically assessed inflammatory markers are highly expressed in adipose tissue, liver and skeletal muscle in obese individuals undergoing bariatric surgery. This obesity-related systemic inflammation is highly correlated to MASH and IMLC.⁴⁹ In this population, inflammatory infiltration is also significantly higher in visceral adipose tissue than in subcutaneous adipose tissue, confirming that inflammation in adipose tissue has a different impact on MASLD depending on its location.⁴⁹ However, the most relevant observation reported on that population is the strong correlation between muscle inflammation, EMLC and liver fibrosis, highlighting muscle inflammation as a potential driver of liver fibrosis.⁴⁹

These observations reinforce the well-described pro-inflammatory state occurring in obesity secondary to an imbalance between pro and anti-inflammatory cytokines. The resulting chronic low-grade inflammation occurs in several insulin-sensitive tissues such as adipose tissue, liver or skeletal muscle and eventually contributes to IR.¹⁵¹ However, these data support the notion that inflammation is a driver of myosteatosis in MASLD and MASLD progression independently of obesity itself.

The particular case of IL15: A mediator expressed by several organs

IL15 is a member of the γc cytokine family expressed by several tissues including the liver.¹⁵² It is therefore difficult to identify its main source in the overall proinflammatory metabolic context (Fig. 5). It is a pro-inflammatory cytokine acting on T lymphocytes.^153,154 Serum levels of IL15 are also increased in MASLD¹⁵⁵ and promote liver steatosis by inducing PPARγ expression, eventually impairing mitochondrial function in liver and adipose tissue.^155,156 It also increases liver inflammatory infiltration by chemotaxis.¹⁵⁵ IL15 is also highly expressed in skeletal muscle and is often defined as a myokine.^157,158 Furthermore, it plays a dose-dependent role in muscle glucose and lipid metabolism.[159], [160], [161]

Indeed, supraphysiological doses of IL15 enhance insulin sensitivity and decrease adiposity in mice fed a HFD.[159], [160], [161] Hence, myosteatosis appears as an inflammatory process that is not only driven by an excessive serum level of FFAs secondary to MASLD but also by pro- and non-inflammatory mediators that alter both skeletal muscle and liver metabolism (Fig. 5).

Conclusion

MASLD is the most common chronic hepatopathy worldwide with only weight loss and physical exercise as proven effective therapeutic weapons in terms of liver fibrosis regression and hence clinical outcomes. Among extrahepatic complications of MASLD, myosteatosis is highly prevalent even in non-cirrhotic stages. This association is compatible with a pathogenic axis between the muscle and liver, the direction of which remains unanswered.

In terms of pathophysiology, only a few mechanisms and mediators have been reported to be involved in the pathogenesis of myosteatosis in MASLD (Fig. 1). However, many others might also be involved via modulation of glucose and lipid metabolism in both the liver and skeletal muscle.

In terms of potential complications, myosteatosis is reported to decrease insulin sensitivity and muscle strength, and to correlate with MASH (Fig. 1). Therefore, myosteatosis could become an interesting screening tool for MASH in clinical practice. However, it is paramount to acknowledge that the causal link between IR, MASH, muscle function decay and myosteatosis has not been well demonstrated, nor has the respective contribution of inter and intramuscular fat to these complications. It is important to consider that the association between MASH and myosteatosis was only reported in obese individuals with myosteatosis semi-quantitatively assessed by CT (Table 3). Studies in other population groups, including all MASLD disease stages and using MRI and MRS techniques to characterise the muscle compartment will surely shed light on the complex muscle-liver-adipose tissue axis that is anticipated to be the main determinant of MASLD pathogenesis and evolution. With these results, it will then be interesting to assess the extent to which specific treatment of myosteatosis by dietary changes, weight loss or targeted therapy, could have an impact on liver disease.

Financial support

This work was supported by a concerted research action grant “Myo-MAFLD” 22/27-122. NL has a mandate of Clinical Researcher from the Fonds de la Recherche Scientifique (FNRS), Belgium.

Authors’ contributions

GH and NL conceived and designed the analysis; GH collected the data, performed the analysis, wrote the first draft of the manuscript; AL and IAL contributed for intellectual content, revised and edited the manuscript; IAL and NL obtained funding; NL supervised the work; all authors revised and accepted the final version.

Conflict of interest

GH received a travel grant from Gilead Sciences. IAL has patents planned, issued, or pending (PCT/EP2022/065769 Ref WO/2022/258788); received research grants from FNRS, Région wallonne and Televie. NL received speaker fees from Gilead Sciences and Fresenius Kabi; received travel grants from Abbvie, Gilead Sciences and Norgine and receives grants from Gilead Sciences. AL has no conflict of interest to declare.

Please refer to the accompanying ICMJE disclosure forms for further details.

Supplementary data

The following are the supplementary data to this article: