BACKGROUND: Previous methods to measure glycohemoglobin (GHb) have been time-consuming or imprecise; we therefore developed a new enzymatic assay for GHb. METHODS: Blood cells were first hemolyzed, and hemoglobin was digested with protease to yield fructosyl amino acid. Fructosyl amino acid oxidase acts on the fructosyl amino acid and generates hydrogen peroxide, which reacts with chromogens in the presence of peroxidase. Total hemoglobin was measured spectrometrically in the same reaction tube. The results were reported as the ratio of the concentrations of GHb and hemoglobin. RESULTS: The measured values were comparable to those determined with a HPLC method and with an immunoassay in blood samples from 2854 patients with diabetes. Regression analysis for the enzymatic assay (y) vs the HPLC method (x) produced the following: r = 0.979; slope, 0.994 [95% confidence interval (CI), 0.986-1.001]; y-intercept, 0.04% (95% CI, -0.09% to 0.01%); n = 2854. For the enzymatic assay (y) vs the immunoassay (x), the regression statistics were as follows: r = 0.982; slope, 1.002 (95% CI, 0.995-1.009); y-intercept, 0% (95% CI, -0.05% to 0.05%); n = 2854. CONCLUSIONS: The values measured by the new enzymatic assay are sufficiently correlated with those of the conventional HPLC method and immunoassay, but the proposed assay for GHb is rapid and has high precision.
BACKGROUND: Previous methods to measure glycohemoglobin (GHb) have been time-consuming or imprecise; we therefore developed a new enzymatic assay for GHb. METHODS: Blood cells were first hemolyzed, and hemoglobin was digested with protease to yield fructosyl amino acid. Fructosyl amino acid oxidase acts on the fructosyl amino acid and generates hydrogen peroxide, which reacts with chromogens in the presence of peroxidase. Total hemoglobin was measured spectrometrically in the same reaction tube. The results were reported as the ratio of the concentrations of GHb and hemoglobin. RESULTS: The measured values were comparable to those determined with a HPLC method and with an immunoassay in blood samples from 2854 patients with diabetes. Regression analysis for the enzymatic assay (y) vs the HPLC method (x) produced the following: r = 0.979; slope, 0.994 [95% confidence interval (CI), 0.986-1.001]; y-intercept, 0.04% (95% CI, -0.09% to 0.01%); n = 2854. For the enzymatic assay (y) vs the immunoassay (x), the regression statistics were as follows: r = 0.982; slope, 1.002 (95% CI, 0.995-1.009); y-intercept, 0% (95% CI, -0.05% to 0.05%); n = 2854. CONCLUSIONS: The values measured by the new enzymatic assay are sufficiently correlated with those of the conventional HPLC method and immunoassay, but the proposed assay for GHb is rapid and has high precision.