Failure of common glycation assays to detect glycation by fructose.

Serum albumin was modified by in vitro glycation with either fructose or glucose, to see whether the common clinical assays for glycation were able to detect both fructose- and glucose-induced changes in protein structure in diabetes. Although fluorescence measurements showed that fructose causes far more protein damage than glucose, neither serum fructosamine (SFA) nor phenylboronate affinity (PBA) glycation assays reflected these changes. The SFA method implied that fructose causes only about 5% of the glycation induced by glucose; with PBA the proportion was 25%. The thiobarbituric acid- and periodate-based assays also greatly underestimated the true extent of fructation. We discuss these discrepancies with respect to the underlying chemistry, emphasizing the difference between aldehydic and ketonic Amadori products (exemplified by fructose and glucose derivatives, respectively). The implications for detecting fructose-induced secondary diabetic complications are also discussed.