HbA1c Reconsidered: Glycated Haemoglobin in the Era of Dynamic Erythrocyte Glucose Metabolism

Glycated haemoglobin (HbA1c) has served for four decades as the standard biomarker for long-term glycaemic control, with global use approaching one billion measurements annually. The biomarker rests on a tacit assumption: that the rate of non-enzymatic haemoglobin glycation is determined chiefly by mean plasma glucose concentration, with erythrocyte intracellular glucose acting as a passive equilibrium with plasma. This assumption requires erythrocyte glucose uptake to be constant—that is, glucose transporter 1 (GLUT1) expression should not vary substantially with physiological state. The 2026 demonstration by Martí-Mateos and colleagues that chronic hypoxia upregulates erythrocyte GLUT1 approximately twofold and per-cell glucose uptake approximately threefold, together with growing evidence that the band 3 N-terminus operates as a bidirectional metabolic switch responsive to haemoglobin oxygenation state, invalidates the constancy assumption. We argue that HbA1c should be reconceived not as a record of extracellular glucose exposure but as the time-integral of erythrocyte intracellular glucose exposure—a quantity that diverges from plasma glucose in a directionally predictable manner under conditions of altered erythrocyte oxygen environment, chronic inflammation, or accelerated red cell turnover. We synthesise existing evidence showing HbA1c underestimates true glycaemia in high-altitude populations (where glycated albumin moves in the opposite direction, confirming the dilution-and-flux mechanism), with parallel implications for obstructive sleep apnoea, cyanotic congenital heart disease, intensive care unit hyperoxia exposure, and chronic kidney disease. We propose four testable predictions, outline a Red Cell Hypoxic Metabolic Index (RHMI) as a candidate correction factor, and argue that the next generation of diabetes monitoring should integrate HbA1c with glycated albumin, continuous glucose monitoring, and erythrocyte metabolic phenotyping in a population-specific manner. The framework does not displace HbA1c but provides the missing physiological lens required to read it correctly in 2026 and beyond.