Glycation and AGEs: real ageing chemistry, oversold as a diet you can hack

Advanced glycation end products, or AGEs, are what forms when a sugar molecule bonds to a protein or fat without an enzyme directing it — the same browning chemistry that colours a seared steak, running slowly inside the body [s1]. They accumulate with age, form faster when blood sugar is high, and stiffen the tissues they attach to, which makes them a genuine part of how bodies age and a real driver of harm in diabetes [s1]. What the evidence does not support is the marketing leap: that eating a "low-AGE" diet or taking an "anti-glycation" supplement measurably slows ageing. The human trials show modest shifts in metabolic markers and say nothing about lifespan [s2].

The chemistry is real and well understood

Glucose and other sugars react non-enzymatically with proteins, lipids and DNA in a slow cascade — the Maillard reaction — that ends in stable AGEs [s1]. The most familiar everyday example is HbA1c, the standard diabetes blood test, which is simply glycated haemoglobin: sugar stuck to a blood protein in proportion to how high glucose has run [s1]. Once formed on longer-lived proteins, AGEs cross-link structural molecules such as collagen, leaving skin, tendons and artery walls stiffer and less elastic; and they bind a cell-surface receptor (RAGE) that switches on inflammation and oxidative stress [s1].

That two-part action — mechanical stiffening plus a pro-inflammatory signal — is why AGEs are implicated in the vascular complications of diabetes, in arterial stiffening and in skin ageing, and it connects glycation to the broader story of chronic low-grade inflammation in ageing [s1]. Bodies generate AGEs internally at a rate that climbs with blood glucose, and also absorb them from food, especially items cooked at high, dry heat [s1].

Cutting dietary AGEs: small, short, surrogate

So does eating fewer AGEs help? The best synthesis is a 2021 meta-analysis of 13 randomised controlled trials comparing low- and high-AGE diets [s2]. It found real but narrow benefits. The low-AGE diets lowered insulin resistance (weighted mean difference −1.204; 95% CI −2.057 to −0.358), fasting insulin (−5.472 μU/mL; −9.718 to −1.234), total cholesterol (−5.486 mg/dL; −10.222 to −0.747) and LDL cholesterol (−6.263 mg/dL; −11.659 to −0.866) [s2]. The other components of the metabolic syndrome did not move [s2].

The limits are the point. Every one of those outcomes is a surrogate marker measured over weeks — not a disease, a death or a year of life — and the trials are small, short and heterogeneous, with none designed to test whether eating this way makes anyone live longer [s2]. A biomarker moving in the right direction is a reason to keep studying an idea, not proof that it works.

Accumulated AGEs track outcomes — but that is not a diet claim

There is firmer evidence that how many AGEs a body has built up is a marker worth watching, even if it does not follow that you can eat your way out of it. In the Dutch Lifelines cohort, 72,880 adults free of diabetes and cardiovascular disease had skin autofluorescence measured — a non-invasive proxy for tissue AGE accumulation [s3]. Over a median of four years, higher readings predicted who developed type 2 diabetes (1,056 people, 1.4%), cardiovascular disease (1,258, 1.7%) and death (928, 1.3%), independent of the metabolic syndrome, glucose and HbA1c [s3].

That the accumulated AGE load tracks hard outcomes fits the biology — but the caveats belong on the same line as the finding. The study is observational, so it shows association rather than cause; the device reads a proxy rather than AGEs directly; and two of its authors are founders and shareholders in the company that makes the reader, a conflict worth holding in view when weighing how useful they judge the tool [s3].

The split verdict

Glycation is real, AGEs accumulate with age and predict disease, and keeping blood glucose in a healthy range — which limits the body's own AGE production — is well worth doing on its own merits [s1][s3]. But the specific consumer claim, that hunting down dietary AGEs or buying anti-glycation skincare and supplements will slow ageing, rests on short trials of surrogate markers, not on any longevity outcome [s2]. In practice a low-AGE diet mostly means less heavily browned, high-heat-cooked and ultra-processed food and more of the dietary patterns already tied to health — so whatever benefit it carries may owe as much to what those diets contain as to the AGEs they leave out [s1][s2].