A-063 Correlation of 1,5-anhydroglucitol with A1c: understanding this relationship in terms of an established physiological model for AG mass balance as a function of

D Stickle,B N Delikkaya,P B Kyle

DOI: https://doi.org/10.1093/clinchem/hvae106.062

IF: 12.114

2024-10-27

Clinical Chemistry

Abstract:BackgroundDiet-derived 1,5-anhydroglucitol (AG) is an unmetabolizable glucose analogue present in plasma (reference interval: 7.4-32.5 μg/mL). In diabetes, hyperglycemia can accelerate AG excretion, producing an inverse relationship between A1c and AG. There are no established guidelines regarding use or interpretation of AG in diabetes, but it is often described as being useful for detection of short-term variation in [glucose]. We examined AG usage at our institution, and, additionally, analyzed the information content of AG as derived from an established model for the relationship between AG and [glucose].MethodsMedical records were scanned for all combined orders for AG and A1c between September 2020 and August 2023. Estimated average glucose (eAG) was calculated from A1c [PMID: 18540046]. AG vs. A1c data were compared to calculations for AG vs. eAG using an established physiological model for AG mass balance [PMID: 9357814].ResultsAmong 104 AG measurements, there were 93 AG/A1c measurement pairs (see Figure). 70% had elevated A1c (n=65). 43% had decreased AG (n=40), all with elevated A1c. Among all AG/A1c pairs, 66% (closed points; n=61) were within the boundaries of model calculations for AG when assuming a simple steady-state [glucose] equal to eAG, a normal glomerular filtration rate, and a normal AG ingestion rate. AG outside of these boundaries (34%; open points, n=32) were almost entirely above the upper boundary (32%; n=30), where results are consistent either with increased AG ingestion rate, decreased glomerular filtration rate, or a discrepancy between eAG and A1c. Only two points (2.2%) were compatible with interpretation of variation in prevailing glucose from eAG.ConclusionsFor our data, paired AG/A1c measurements were in majority well characterized by a physiological model based simply on steady-state (invariant) [glucose] equal to eAG as derived from A1c alone. Only a small fraction of AG results suggested possible variation in [glucose].

medical laboratory technology