These are the physiological effects of infusing 100ml of concentrated (20%) human albumin into a patient. This is 20% albumin, a sticky protein concentrate with a variable amount (48-100mmol/L) of sodium in it.
This infusion contributes only 100ml of water. However, more water is attracted into the intravascular space by the sudden rise in plasma oncotic pressure. The movement of water in this situation is determined primarily by capillary permeability, and its influence on the transcapillary escape of albumin molecules.
To calculate the magnitude of intravascular volume expansion in this situation is more complicated than merely working out the distribution of fluid and electrolytes. Instead, I turn to raw data.
Following the early pioneers, in 1976 post-op patients were infused with 50g of albumin in varying concentrations. The volume expansion for all these patients was about 500ml, working out to be 11ml of volume expansion per 1g of "retained" albumin (i.e. accounting for some "escaped" albumin). A similar study had later confirmed these findings in a group of bypass patients, arriving at a figure of 7.7ml volume expansion per 1g of albumin. Additionally, a more recent paper reports that 200ml of 20% albumin results in a 430ml volume expansion among septic patients (is that 10.75 ml per gram)- with the expansion maximal at 30 minutes post infusion, and petering out rapidly afterwards (suggesting that in septic ICU patients the transcapillary escape rate is increased by as much as 200%, to 10% per hour).
Yes, this does mean that 4% albumin does not expand volume as much as you would expect. Consider: 4% albumin is iso-oncotic with the blood. It would not attract any extra water from the interstitial space. So the volume expansion you would expect would be at maximum 500ml (i.e. the same volume as the albumin solution). However, this solution is really just normal saline with 20g of albumin added to it. The saline, left to its own devices, would distribute 75% of its volume to the interstitial space, leaving behind only about 125 ml of intravascular volume expansion. At least the albumin keeps some oncotic pressure in that space, preventing some of this escape. That oncotic pressure is enough to retain about 7-11 ml of water per 1g of albumin in the intravascular space, so the 500ml bolus of 4% albumin will only expand the circulating volume by about 140-220 ml at best. And there is some data to support this: Statkevicius et al (2019) gave 12ml/kg of 5% albumin (so about 840ml per 70kg patient) and got only about 6.5-7.5ml/kg plasma volume expansion, which is roughly what is predicted by the abovementioned values.
In short, we can approximate that in our healthy cylinder model, the 20g of albumin will attract about 220ml of water into the intravascular space.
Let us not become distracted from the oncotic marvels of albumin by the petty movements of 10mmol of sodium.
Instead, let us observe what happens to serum osmolality when 220ml of free water suddenly feels compelled to stay in the intravascular space.
The intravascular compartment volume increases by 220ml – from 5000ml to 5220ml. The increase in intravascular volume is 4.4% - outside the volume receptor sensitivity threshold.
This is explained in detail elsewhere.
The water dilutes the intravascular space. In this simplified model, only water migrates into the vessels. The serum osmolality should drop by 16.1mOsm/L (5.75%) and this would be sensed by osmoreceptors. There is a decrease in the release of antidiuretic hormone, and diuresis ensues, correcting the hypoosmolar state.
But does this actually happen experimentally?
The suggestion that concentrated albumin causes a dilution of the circulating volume and produces diuresis by decreasing vasopressin secretion is not very well supported.
In reality, albumin alone has a negligible diuretic effect. There is a tendency to co-administer it with frusemide, hoping for an enhancement of fluid removal from the interstitial space, but this practice isnot supported by the evidence either.
In fact, the notion that albumin reduces serum osmolality by such a massive amount is also suspect. A case series of 3 cirrhosis patients reports only insignificant changes in serum osmolality following albumin infusion.
How can this be?
If we observe the Gibbs-Donnan effect interactions in the capillary, we will notice that as plasma oncotic pressure rises (and drags water osmotically out of the interstitial compartment), there is an increase in the net anionic charge of the plasma. This attracts positively charged ions like sodium out of the interstitial fluid and into the plasma. The charge effect is powerful. Theodore J. Peters reports that the charge of albumin at physiologic pH is something like -17 to -19. This very feature keeps it from passing through the charge barriers at the glomerulus.
This is supported by evidence, at least when it is studied in patients with cirrhosis. Those who undergo paracentesis and receive no albumin end up becoming hyponatremic; but in those who receive albumin the serum sodium remains unchanged.
Thus, the measured osmolality of the plasma after albumin infusion may not actually change very much- enough sodium enters the intravascular space to keep the osmolality stable.
The distribution and metabolic fate of albumin is covered elsewhere.