Effect of insulin on systemic and renal handling of albumin in nondiabetic and NIDDM subjects

Insulin resistance and hyperinsulinemia cluster with microalbuminuria in both diabetic and nondiabetic subjects, but the mechanism underlying this association is unknown. To test the hypothesis that insulin influences protein permeability, we measured the albumin transcapillary escape rate (TER) by the I-131-labeled albumin technique in 12 healthy volunteers and 12 normoalbuminuric NIDDM patients (fasting plasma glucose, 10.9 +/- 1.3 mmol/l) during 4 h of isoglycemia with high (1.1 mU . min(-1) . kg(-1)) or, on a different day, low (0.1 mU . min(-1) . kg(-1)) insulin infusion. In both patients and control subjects, high insulin was associated with a 7% decrease in blood volume (P = 0.006) and a 6% decrease in diastolic blood pressure (P < 0.02), these two changes being related to one another (r = 0.56, P < 0.01). Basal albumin TER was similar in patients (8.4 +/- 0.5% . h(-1)) and control subjects (7.7 +/- 0.7% . h(-1)) and was not significantly changed by high insulin in either group (patients vs. control subjects, 7.3 +/- 0.9 vs. 6.2 +/- 0.4% . h(-1); NS vs. low insulin). In contrast, high insulin increased renal albumin excretion (from 3.6 +/- 0.8 to 5.4 +/- 1.1 pg/min, P < 0.01) and clearance rate (0.09 +/- 0.02 to 0.13 +/- 0.03 mu l/min, P < 0.001) in patients but not in control subjects. To localize the effect of insulin along the nephron, we measured the urinary excretion of N-acetyl-beta-D-glucosaminidase (beta-NAG), released by the proximal tubule; retinol-binding protein (RBP), reabsorbed by the proximal tubule; and Tamm-Horsfall protein (THP) and epidermal growth factor (EGF), both secreted by the distal tubule. For both beta-NAG and RBP, but not EGF or THP, insulin enhanced urinary excretion (diabetics vs. controls: beta-NAG, 0.48 vs. -0.15 mu U/min [P = 0.03]; BBP, 78 vs. -32 ng/min [P = 0.05]). In conclusion, physiological hyperinsulinemia does not affect systemic albumin permeability in healthy subjects or normoalbuminuric NIDDM patients. In contrast, in NIDDM patients, but not in healthy subjects, insulin increases the urinary excretion of albumin and protein markers of proximal tubular function. The significance of this finding for the pathogenesis of diabetic nephropathy remains to be established.