The permeability of the microvasculature was studied for the first 50 min after an intravenous injection of between 6 and 10 U of crystalline insulin in seven juvenile diabetics whose mean age was 22 yr. No patient had clinical signs of neuropathy or microangiopathy. The following variables were determined before and after the insulin injection: plasma volume, PV (125I-labeled and 131I-labeled human serum albumin), plasma protein concentration, hematocrit, blood pressure, and pulse rate. PV was measured 40, 45, and 50 min after the insulin injection. Transcapillary escape rate of albumin (TER), defined as the fraction of intravascular mass of albumin that passes to the extravascular space per unit of time, was determined after insulin from the disappearance of the intravenously injected 125I-labeled human serum albumin. The mean blood glucose concentration decreased from 15.1 to 10.8 mmol/L at 50 min after the insulin injection. Blood pressure remained unchanged, while a significant increase in pulse rate occurred after the insulin administration. A reduction in PV, calculated after a mixing period lasting 10 (−3.9%), 15 (−2.3%), and 20 (−1.7%) min, was found after the insulin injection. The last-mentioned difference is not significant statistically. The venous hematocrit remained unchanged; this variable was measured so accurately that a 1% decrease in plasma volume could be excluded with 99% confidence. Plasma protein concentration also remained unchanged. Except the one patient with an extremely high TER (10.6%/h), the remaining six patients had normal TER values after the insulin injection (mean, 5.8; range, 4.5–7.2%/h) compared with our previously investigated control group (mean, 5.4; range, 3.8–7.2%/h) and a comparable group of six short-term juvenile diabetics investigated during good (mean, 5.8; range, 4.4–7.2%/h) and poor (mean, 6.9; range, 5.7–7.7%/h) metabolic regulation (these patients did not receive insulin before the TER determination). In agreement with a previous study, thus, we found a significant reduction in plasma volume and in intravascular mass of protein after a small dose of crystalline insulin intravenously. But the present finding of an unchanged hematocrit and a normal TER does not support the suggestion made in the previous study, viz. that there is an increased transfer of fluid and albumin out of the vascular system. The most likely explanation for the abovementioned reductions is bad mixing (pooling), probably a result of the insulin-induced increase in adrenergic nervous activity, causing vasoconstriction and reduced peripheral blood flow, as demonstrated previously.

This content is only available via PDF.