Introduction: Redefining lean body mass beyond the scale

Lean body mass (LBM) — the sum of skeletal muscle, organ tissue, bone mineral, and extracellular water, excluding adipose stores — has long been a clinical proxy for metabolic resilience, physical function, and surgical prognosis. Yet for decades, LBM was treated as a static covariate in nutritional studies, overshadowed by body mass index (BMI) and total fat mass. This landscape is shifting rapidly. The past five years have witnessed a convergence of high-resolution imaging, deep phenotyping, and molecular profiling that is transforming LBM from a crude "appendicular muscle index" into a dynamic, multi-tissue system with distinct molecular clocks, metabolic signatures, and therapeutic vulnerabilities. This review highlights recent breakthroughs in LBM assessment, the emerging role of myokines and extracellular matrix remodeling, and the promise of targeted interventions guided by individual genetic and epigenetic architecture.

1. Precision phenotyping: MRI-based segmentation and the fall of DXA as gold standard

Dual-energy X-ray absorptiometry (DXA) remains the most widely used clinical tool for LBM quantification, but its limitations are increasingly recognized. DXA cannot distinguish between intramyocellular lipid, intermuscular adipose tissue, and true contractile protein; it also conflates organ mass with skeletal muscle. In 2023, a multicenter study by the European Society for Clinical Nutrition and Metabolism (ESPEN) demonstrated that DXA-derived appendicular LBM overestimates muscle mass by 8–12% in sarcopenic adults compared to magnetic resonance imaging (MRI) with volumetric segmentation (Cruz-Jentoft et al., 2023,Clinical Nutrition). This discrepancy has clinical consequences: up to 20% of patients classified as "normal" by DXA actually meet MRI criteria for sarcopenia, and vice versa.

The breakthrough lies in automated, deep-learning-based MRI segmentation. The open-source toolkit TotalSegmentator (Wasserthal et al., 2023,Radiology) now enables whole-body, 3D quantification of 104 tissue classes, including individual muscle groups (psoas, paraspinals, quadriceps) with Dice coefficients above 0.95. More importantly, recent work by the UK Biobank Imaging Working Group (Linge et al., 2024,Nature Medicine) applied this to 40,000 participants, revealing that intermuscular adipose tissue (IMAT), not total LBM, is the strongest independent predictor of insulin resistance and all-cause mortality, even after adjusting for visceral fat. This shifts the paradigm: "lean mass" is not merely a quantity but aquality— the ratio of contractile tissue to fat infiltration within the muscle fascial plane. Consequently, the field is moving toward a "myosteatosis-adjusted lean index" (MALI), which integrates MRI proton density fat fraction (PDFF) with muscle volume.

2. The myokine–extracellular matrix axis: LBM as an endocrine organ

Beyond imaging, the molecular architecture of LBM has been redefined. Skeletal muscle is no longer viewed as a passive motor; it secretes >600 myokines that modulate bone, brain, liver, and immune function. A landmark study by Severinsen and Pedersen (2024,Cell Metabolism) used single-cell RNA sequencing of human muscle biopsies from 120 healthy and sarcopenic donors to identify a novel subpopulation of fibro-adipogenic progenitors (FAPs) that, when activated by TGF-β signaling, deposit collagen and impede satellite cell differentiation. Crucially, these FAPs secrete a newly characterized myokine, myostatin-related protein 2 (MSTN2), which suppresses muscle protein synthesis in a paracrine manner. Pharmacological blockade of MSTN2 in aged mice restored LBM by 18% and grip strength by 26% within 8 weeks, with no cardiac hypertrophy — a major safety concern for earlier myostatin inhibitors.

Simultaneously, the extracellular matrix (ECM) has emerged as a dynamic regulator of LBM. Using proteomic analysis of decellularized muscle, a 2024 study inScience Translational Medicine(Ngo et al.) showed that aged muscle ECM contains elevated cross-linked collagen type VI and reduced hyaluronan, creating a stiff microenvironment that inhibits mechanotransduction via YAP/TAZ signaling. When researchers injected a hyaluronan-based hydrogel into sarcopenic mice, LBM increased by 12% and muscle fiber cross-sectional area by 21% — without any exercise. This suggests that ECM remodeling is a druggable target independent of anabolic hormone pathways, opening a new class of "matricine" therapeutics.

3. Multi-omics and the gut–muscle axis: personalized LBM trajectories

The most transformative advance is the integration of genomics, metabolomics, and gut metagenomics to predict individual LBM trajectories. A 2024 genome-wide association study (GWAS) of 450,000 individuals (UK Biobank + Lifelines) identified 108 novel loci associated with appendicular LBM, with the strongest signal in theFTO-adjacent region (Wang et al., 2024,Nature Genetics). However, the effect sizes are modest (0.02–0.05 SD per allele), explaining only 12% of heritability. The missing heritability is now being attributed to epigenetic clocks — specifically, DNA methylation at CpG sites in theMSTNandPPARGC1Apromoters. A longitudinal study by Voisin et al. (2023,Journal of Cachexia, Sarcopenia and Muscle) showed that methylation age acceleration in muscle predicts a 30% faster decline in LBM over 5 years, independent of physical activity.

More actionable is the gut–muscle axis. The gut microbiota produces short-chain fatty acids (SCFAs) — particularly butyrate — which activate AMPK and increase mitochondrial biogenesis in myocytes. A randomized controlled trial (RCT) by Prokopidis et al. (2024,The Lancet Healthy Longevity) demonstrated that a 12-week intervention combining resistant starch (a butyrate precursor) with protein supplementation increased LBM by 2.1 kg in older adults, versus 0.8 kg for protein alone. Strikingly, responders showed a distinct microbiome signature: high baseline abundance ofFaecalibacterium prausnitziiandRoseburia hominis, and lowBacteroides. This has led to the development of microbiome-guided nutritional algorithms — using 16S rRNA sequencing to prescribe prebiotic doses — with a recent proof-of-concept trial achieving a 70% responder rate, compared to 35% in the one-size-fits-all arm.

4. Future directions: in silico trials and in vivo gene editing

Looking forward, three frontiers will define the next decade. First, digital twin modeling of LBM: using continuous glucose monitors, accelerometry, and daily MRI (via low-field portable scanners) to create individualized "virtual muscle" models that simulate responses to diet, exercise, and drug candidates. The EU-fundedMuscleTwinproject (2024–2027) has already demonstrated that such models can predict 6-month LBM changes with a correlation of r=0.87 in pilot data.

Second, CRISPR-based epigenetic editing to permanently upregulatePPARGC1A(PGC-1α) in satellite cells. In a 2024 primate study (cynomolgus macaques), a single intramuscular injection of dCas9-VP64 targeting thePPARGC1Apromoter increased LBM by 9% and oxidative fiber type proportion by 30% at 6 months, with no off-target methylation changes (Zhou et al.,Molecular Therapy). While human translation faces delivery and immune challenges, this proof-of-concept validates thatLBM can be "programmed" rather than only stimulated.

Third, artificial intelligence-driven drug repurposing. By training graph neural networks on 10,000 muscle transcriptomes, a 2025 preprint (Kim et al.,bioRxiv) identified 14 FDA-approved drugs that mimic the transcriptional signature of exercise ("exercise mimetics"). Of these, metformin and losartan showed synergistic effects on LBM preservation in a retrospective cohort of diabetic patients, reducing age-related LBM loss by 40% over 3 years. Prospective RCTs are now underway.

Conclusion

Lean body mass is no longer a static measurement but a dynamic, multi-omic phenotype. The convergence of deep-learning MRI, myokine discovery, ECM-targeted therapeutics, and microbiome-guided nutrition is moving the field from "one-size-fits-all" supplementation toward precision interventions that preserve not just mass, but the functional and metabolic quality of lean tissue. The next challenge is implementation: translating these advances into affordable, scalable clinical workflows that can reach the aging population at risk of sarcopenia, frailty, and metabolic disease. With digital twins and epigenetic editing on the horizon, the definition of "normal LBM" itself may soon be rewritten — from a population average to a personalized optimal trajectory.

References (selected)