Two of your patients sit at an HbA1c of 7.4%. One has been there for a decade with clean retinal exams and a normal urine albumin-to-creatinine ratio. The other, matched for duration and roughly matched for everything else, is accumulating background retinopathy and microalbuminuria. Same number. Different trajectory. Part of the answer is something the HbA1c never showed you.
HbA1c measures a stable end-product: glucose bound slowly and reversibly to hemoglobin. It is an excellent integrated marker of average glucose. But it is a poor witness to a faster, more destructive chemistry happening in parallel the chemistry of the reactive 1,2-dicarbonyls, above all methylglyoxal (MGO).
MGO is not a glucose molecule. It is a small, highly reactive carbonyl formed continuously as a spillover product of glycolysis, mainly from the triose phosphates. It is orders of magnitude more reactive than glucose itself, and it does not wait around: it modifies arginine and lysine residues on long-lived proteins, forming adducts such as MG-H1 and, downstream, advanced glycation end-products (AGEs) that engage RAGE signaling and drive microvascular injury in exactly the tissues we worry about kidney, retina, vasculature, nerve.
Here is the part worth internalizing. The body has a dedicated, high-capacity system for clearing MGO before it can do this damage: the glyoxalase system glyoxalase 1 and 2 (GLO1/GLO2), working with reduced glutathione (GSH) as an obligate cofactor. Tissue exposure to MGO is therefore not just a function of how much is generated (which rises with hyperglycemia); it is a function of the balance between generation and clearance.
And clearance is where diabetes deals a second, quieter blow. In obesity, insulin resistance, and type 2 diabetes, GLO1 expression and activity are frequently reduced at the same time MGO generation is increased. Generation up, clearance down a double hit that HbA1c cannot see, because HbA1c is measuring glucose, not the dicarbonyl balance.
The honest version. This is a lens, not a lab test you can order Monday. Plasma MGO and urinary MG-H1 are research-grade measurements, not point-of-care tools. Skin autofluorescence offers a non-invasive, integrated readout of tissue AGE accumulation and is the closest accessible proxy, but it is not yet standard care. So the clinical value here is conceptual: it explains, mechanistically, why complication risk varies among patients at matched glycemic control, and it reframes “glycemic control” as involving clearance capacity, not just the glucose number.
Clinical takeaway. Think of glycation as a balance sheet: dicarbonyl generation on one side, glyoxalase clearance on the other. HbA1c reports one input. When a well-controlled patient’s complications don’t match their number, the dicarbonyl side of the ledger is one of the reasons the arithmetic doesn’t feel like it adds up.