Towards personalised protein prescription in critically ill patients during metabolic stress and recovery: a narrative review and proposed future research directions

Protein provision in critically ill patients has traditionally been guided by population-based, body weight-derived targets and the assumption that higher delivery may attenuate catabolism, muscle wasting and functional decline. Recent large, randomised trials comparing higher versus lower protein delivery have not demonstrated benefit on their primary patient-centred outcomes. Renal subgroup analyses have raised additional safety concerns, but these findings do not establish that patients without acute kidney injury (AKI) benefit from higher protein exposure. This narrative review critically appraises randomised trials, systematic reviews and meta-analyses, together with selected mechanistic and non-randomised evidence, and proposes future research directions for personalised protein prescription throughout critical illness and recovery. The capacity to receive and metabolically handle administered protein is unlikely to remain constant, such that protein dose should be prescribed according to metabolic phase and direction of clinical trajectory rather than intensive care unit (ICU) day alone. Conventional weight-based dosing may be imprecise because patients with similar body weight can differ substantially in fat-free mass, skeletal muscle mass and adiposity. Computed tomography (CT), muscle ultrasound and bioelectrical impedance analysis (BIA) may identify low muscularity or structural change, but require careful interpretation because of limitations related to timing, fluid status, methodology and measurement variability; none has been validated to determine an individual protein requirement. Monitoring should distinguish prescribed protein targets, actual protein delivery, gastrointestinal capacity to receive enteral nutrition, organ constraints and metabolic response. Personalised protein prescription should not be regarded as an alternative to evidence generated by adequately designed randomised controlled trials (RCTs). Future research should prioritise phase-specific and phenotype-enriched RCTs, prospective testing of body-composition- and response-informed strategies, and outcomes reflecting muscle health, physical function, independence and quality of life while appropriately accounting for death. A phase-, phenotype- and response-informed approach is biologically plausible and testable but is not yet a validated dosing algorithm.