Effects of albumin and immunoglobulin A on fructosamine assay.

Serum fructosamine, albumin, and IgA were measured in three groups of human subjects: 54 nondiabetic normal individuals, 149 nondiabetic patients, and 149 diabetic patients. Normal subjects had significantly (P less than 0.05) higher fructosamine (2.91, SD 0.33, mmol/L) than did nondiabetic patients (2.49, SD 0.46, mmol/L). Each of these groups had significantly (P less than 0.05) lower fructosamine than did diabetic patients (3.76, SD 1.16, mmol/L). Increased fructosamine appeared to be associated with increases in both albumin and IgA. However, fructosamine was significantly (P less than 0.05) correlated with neither albumin nor IgA in the normal group, with only albumin in the nondiabetic group, but with both albumin and IgA in the diabetic group. Selective combinations of these populations not only shifted these significances but also eliminated some of the correlations. Our results suggest caution regarding the diagnostic role and universality of fructosamine and of its correlation with IgA as indicated by others.