Probiotic supplementation as a nutritional strategy for the prevention and management of sarcopenia in older adults.

Probiotic users in the UK Biobank cohort had greater handgrip strength and higher fat-free mass compared to non-users.

• Handgrip strength was 27.2 ± 4.8 kg in probiotic users vs. 25.1 ± 5.2 kg in non-users (UK Biobank, n = 5,000)
• Fat-free mass was 57.4 ± 10.6 kg in probiotic users vs. 56.2 ± 11.0 kg in non-users
• Probiotic users also exhibited lower inflammatory burden compared to non-users
• Data were drawn from UK Biobank (n = 5,000), NHANES 2011–2018 (n = 3,500), and GMrepo (n = 800)

Probiotic-conditioned media (PCM) increased myotube fusion index and diameter in C2C12 myotubes in vitro.

• Fusion index was 85.6 ± 2.3% in PCM-treated cells vs. 78.9 ± 3.2% in controls (p < 0.01)
• Myotube diameter was 45.2 ± 6.3 µm in PCM-treated cells vs. 40.1 ± 5.7 µm in controls (p < 0.05)
• C2C12 myotubes were used as the in vitro model system

PCM upregulated myogenic regulatory factors and a mitochondrial biogenesis marker in C2C12 myotubes.

• MyoD expression increased 2.1-fold following PCM treatment
• Myogenin expression increased 1.8-fold following PCM treatment
• Myosin heavy chain (MyHC) expression increased 2.2-fold following PCM treatment
• PGC-1α, a marker of mitochondrial biogenesis, was increased 1.9-fold

PCM reduced pro-inflammatory markers NF-κB and TNF-α in C2C12 myotubes.

• NF-κB was decreased approximately 0.60-fold following PCM treatment
• TNF-α was decreased approximately 0.65-fold following PCM treatment
• These reductions suggest an anti-inflammatory effect mediated through the gut-muscle axis

Molecular docking simulations showed that butyrate had high binding affinity for key metabolic regulatory proteins AMPK, mTOR, and NF-κB.

• Butyrate binding affinity for AMPK was -6.9 kcal/mol
• Butyrate binding affinity for mTOR was -6.4 kcal/mol
• Butyrate binding affinity for NF-κB was -6.0 kcal/mol
• These findings suggest short-chain fatty acids (SCFAs) may mediate probiotic effects on muscle metabolism and inflammation through direct interactions with these regulatory proteins

An XGBoost machine learning model trained on integrated multimodal features achieved 88% accuracy in predicting sarcopenia risk.

• Random Forest and XGBoost models were used for sarcopenia risk prediction
• Models used integrated multimodal features from the combined datasets
• A 70% training and 30% validation split was applied
• XGBoost achieved the highest performance at 88% accuracy