The familiar glucose lowering drug may also have an underappreciated anti-dicarbonyl footprint. The evidence is compelling, but the mechanism deserves an honest caveat.
Enrique C. Fernández, MD
Network Based Diabetes Care
We reach for metformin to lower glucose. That is the label, that is the mechanism we teach, and it is true. But there is a second thing metformin appears to be doing, quietly, off to the side, and largely unmeasured in ordinary practice, that connects the dicarbonyl story to a drug you prescribe every day.
In 1999, Beisswenger and colleagues reported something easy to overlook. Metformin-treated patients with type 2 diabetes had systemic methylglyoxal levels roughly 30% lower than patients on other agents at equivalent glycemic control. Read that again: the difference was independent of HbA1c. Whatever metformin was doing to MGO, it was not simply a downstream consequence of lowering glucose. Later work has continued to link metformin to altered methylglyoxal handling.
Several pathways, working at once
So how does metformin lower a reactive dicarbonyl? This is where the story gets genuinely instructive, because the honest answer is several ways at once, and we cannot cleanly separate them.
01 Enzyme induction
Metformin activates AMPK and the downstream Nrf2 transcriptional program, which includes GLO1 among its targets. It may induce the clearance enzyme itself.
02 Direct trapping
A guanidine group can react directly with methylglyoxal to form a chemical adduct, trapping the molecule without changing enzyme abundance.
03 Reduced generation
Metformin alters glycolytic and gluconeogenic flux and reduces oxidative stress, shifting how much methylglyoxal is generated in the first place.
Here is the teaching point, and it is a bigger one than metformin. It is tempting to compress this into a clean sentence, “metformin works by boosting glyoxalase,” and put it on a slide. That sentence overreaches. GLO1 induction is one contributor among several, and the direct chemical trapping may account for a large share of the observed effect independent of any enzyme change. The mechanism is real; the specific attribution is not settled.
THE HONEST VERSION
No trial has shown that metformin’s complication benefits are caused by its MGO-lowering effect.
That the two coexist is mechanistically satisfying, but coexistence is not causation. This is precisely the kind of plausible but unproven link that deserves a caveat rather than a claim.
What we can say is that metformin has an underappreciated anti-dicarbonyl footprint that fits its reputation for durable, complication-sparing benefit. That footprint is a reasonable part of why metformin remains first-line, without being a proven mechanism of its outcomes.
CLINICAL TAKEAWAY
Appreciate metformin’s dicarbonyl footprint; do not oversell it.
When a patient asks why this old, inexpensive drug is still first-line, “it does more than lower glucose” is a fair and interesting answer, as long as you hold the mechanism honestly: plausible contributor, not established cause.