Assessment of Nutritional Status in Adult Patients with Acute Myeloid Leukemia: The Role of Screening Tools, Composite Indices, Body Composition, and Physical Function—A Structured Narrative Review

Nutritional deterioration in adults with acute myeloid leukemia (AML) may be present at diagnosis and may evolve rapidly during intensive chemotherapy, lower-intensity treatment, and allogeneic hematopoietic cell transplantation (allo-HCT). This structured narrative review critically examines nutritional screening and diagnostic instruments, anthropometry and body composition, laboratory-derived composite indices, and objective physical-function measures. PubMed was searched for publications from 1 January 2015 through 13 July 2026; Google Scholar and reference lists were used foradditional citation searching. Of 292 records screened, 52 full-text reports were assessed and 34 empirical studies were included. Methodological limitations were appraised with the 2018 Mixed Methods Appraisal Tool. Evidence was heterogeneous and predominantly observational. PG-SGA and NRS-2002 identified nutritional risk and nutrition-impact symptoms, but their classifications were not interchangeable with GLIM diagnosis, body-composition phenotyping, or prognostic indices. BMI alone did not capture muscle depletion. CT, quantitative CT, BIA-derived phase angle, skeletal muscle indices, and repeated anthropometry detected clinically relevant changes, although thresholds varied and fluid shifts may bias BIA. Albumin-based indices, including CONUT, sCONUT, NRI, GNRI, PNI, and the C-reactive protein/albumin ratio, were associated with outcomes in several retrospective cohorts, but largely reflect inflammation, disease burden, and treatment effects as well as nutrition; no head-to-head evidence establishes superiority. SPPB, gait speed, walking tests, chair stands, and grip strength provided prognostic information in selected older and transplant cohorts, including patients treated with hypomethylating agents plus venetoclax. The evidence supports repeated, multimodal assessment, while remaining insufficient to prove that any single tool or tool-guided intervention improves survival. Future studies require standardized definitions, sex-stratified validation, contemporary treatment cohorts, and trials integrating nutrition, symptom control, and exercise.