Proteinuria in diabetes mellitus: role of spontaneous and experimental variation of glycemia.

The excretion rates of albumin, IgG and beta 2-microglobulin were studied in insulin-dependent diabetic patients with (Albustix positive) and without (Albustix negative) clinical proteinuria and in a group of nondiabetic controls. In patients negative for clinical proteinuria, the mean excretion rate of albumin and IgG was increased but that of beta 2-microglobulin was normal. HbA1, a measure of excess glycemia, was positively correlated with both albumin and IgG urinary excretion rates. Vigorous correction of glycemic control significantly reduced IgG excretion in nine patients. In patients positive for clinical proteinuria, albumin and IgG clearances were inversely correlated with GFR, the filtration of IgG increasing relatively more than that of albumin as GFR declined. A negative hyperbolic correlation was found between GFR and beta 2-micro-globulin excretion. In this group HbA1 was unrelated to excretion rates or clearances of albumin and IgG. In clinically proteinuric patients, long-term correction of hyperglycemia by continuous subcutaneous insulin infusion failed to check the increasing albumin and IgG filtration. The microproteinuria of diabetes is glomerular in origin, is influenced by prevailing glycemia, and is reversible by vigorous glycemic control. In the clinically proteinuric phase, by contrast, selectivity is progressively lost and, as GFR falls, proteinuria becomes of a mixed glomerular and tubular origin. There is no evident association with prevailing glycemia, and metabolic near-normalization dose not appear to affect progression over the period of observation considered. This study suggests that clinical proteinuria denotes the installation of a self-maintaining process, largely independent of the diabetic metabolic disturbance which gave rise to it. Prevention of clinical diabetic nephropathy by metabolic correction may be achievable only in the phase of early microproteinuria.