Ju Hun Oh, Cheol Ryong Ku, Chang-Myung Oh
Sarcopenia has consensus criteria on three continents and an all-cause mortality hazard ratio near 1.6, yet no drug is approved for the primary age-related condition and agents that reliably enlarge muscle have not improved the strength and gait-speed endpoints that define it. We approach this translational failure from the biology rather than the pipeline. Ageing degrades four properties of skeletal muscle that jointly determine force production and are each druggable in principle: innervation, through motor-unit loss and failed compensatory reinnervation; metabolic competence, through declining mitochondrial content and respiratory capacity; regenerative capacity, through satellite-cell drift into senescence within an inflamed niche; and intramuscular lipid, through fibro-adipogenic progenitor-derived adipogenesis. Mapping every therapeutic class onto these axes shows that anabolic agents engage none of them, that classes engaging a quality axis engage exactly one, and that no sarcopenia programme has been designed around innervation or intramuscular lipid. Two experiments make the innervation gap concrete: soluble activin receptor IIB failed to protect denervated muscle, and follistatin overexpression, added to voluntary exercise, rescued neither motor-unit loss nor neuromuscular junction transmission beyond exercise alone. The clinical record remains compatible with three explanations no completed trial can separate: a mismeasured exposure, since dual-energy X-ray absorptiometry lean mass is not muscle; a mismeasured outcome, since the minimal detectable change of gait speed and the Short Physical Performance Battery exceeds their own consensus thresholds for substantial clinical change; and genuinely uncoupled biology. We appraise 15-PGDH inhibition, now in early clinical development, together with its counter-case, and propose a mass-matched trial design capable of separating the three.