Harris-Benedict (1919) systematically overestimates BMR in modern populations by ~5%. Mifflin-St Jeor (1990) was validated against indirect calorimetry on 498 subjects and remains the most accurate predictive equation absent direct measurement.
The 1.2 through 1.9 multipliers are FAO/WHO/UNU joint-consultation values. They're rough. Real-world expenditure can vary ±15% from prediction. Start here, then adjust based on weekly weight trend.
For body-composition goals, target ≥1.6 g/kg protein. Especially important on a GLP-1 (semaglutide, tirzepatide) — appetite suppression makes under-protein common, accelerating lean-mass loss. See the GLP-1 comparison.
Estimates energy needs. Not a prescription. If you have a clinical condition affecting metabolism — thyroid dysfunction, recovery from disordered eating, diabetes — confirm targets with a clinician.