Abstract / Summary
Abstract Background Exercise and physical activity have become standard parts of frailty care, yet much of the literature treats frailty biology and exercise prescription as separate problems. Guidance is therefore less clear on which component should lead when weakness, breathlessness, gait instability, or post-acute deconditioning dominates the presentation. Main text We treat frailty as a multisystem state of reduced physiologic reserve and use this perspective to set rehabilitation priorities. Evidence from frailty-defined cohorts was preferred. Evidence from sarcopenia trials, broader exercise studies in older adults, and disease-specific cardiopulmonary or post-acute rehabilitation was included where relevant and explicitly identified as extrapolative. Progressive resistance training remains the core intervention when low force output, slow chair rise, or transfer difficulty most directly constrains function. Aerobic or cardiopulmonary training can take precedence when dyspnea, poor walking endurance, or orthostatic intolerance most constrains activity. Balance and motor-cognitive training are appropriate when unstable gait, turning difficulty, or dual-task breakdown limits participation. Because most frail older adults present with mixed deficits, multicomponent exercise serves as the default delivery framework. Low-load strengthening, neuromuscular electrical stimulation, digital tools, and mind–body exercise function as bridges or adjuncts. These priorities integrate physiological reasoning with indirect clinical evidence. Direct trials in frailty-defined populations should now test whether bottleneck-matched prescription improves outcomes compared with usual or non-matched exercise prescription. Conclusions Current evidence supports pathophysiology-informed exercise prioritization as a clinical framework. Formal validation as a treatment-matching algorithm remains the next research step. The practical task is to identify the dominant reserve-limiting bottleneck, begin at a tolerable dose, progress load and complexity according to explicit criteria, and distinguish direct frailty evidence from justified extrapolation.