Insulin is secreted in a high frequency pulsatile manner. These pulses are
delivered directly into the portal vein and then undergo extraction and dil
ution before delivery into the systemic circulation. The reported frequency
of these insulin pulses estimated in peripheral blood varies from an inter
pulse interval of 4-20 min. We postulated that this discrepancy is due to t
he attenuation of the pulse signal in the systemic circulation vs. the port
al circulation. In the present study we measured pulsatile insulin release
directly in the portal circulation of human subjects who had indwelling tra
nsjugular intrahepatic portasystemic stent shunts (TIPSS) to decompress por
tal hypertension. We quantitated pulsatile insulin secretion in both the ov
ernight fasted state (fasting) and during a hyperglycemic clamp (8 mmol/L).
Direct portal vein sampling established that pulsatile insulin secretion i
n humans has an interval (periodicity) of approximately 5 min. The amplitud
e (and mass) of the insulin concentration oscillations observed in the port
al vein was approximately 5-fold greater than that observed in the arterial
ized vein and was similar to that observed in the dog. Increased insulin re
lease during hyperglycemia was achieved through amplification of the insuli
n pulse mass. In conclusion, direct portal vein sampling in humans revealed
that the interpulse interval of insulin pulses in humans is about 5 min, a
nd this frequency is also observed when sampling from the systemic circulat
ion using a highly specific insulin assay and 1-min sampling, but is about
4-fold greater than the frequency observed at this site using single site R
IAs. We confirm that enhanced insulin release in response to hyperglycemia
is achieved by amplification of these high frequency pulses.