A brief primer on intravenous milk

vivianimbriotis | Oct. 24, 2022, 2:08 p.m.

Disclaimer: do not infuse milk into anyone.

Having recently injected milk into the veins of a patient about to die, who survived for some hours after the operation, and believing that this therapeutic measure is one of great value, as a substitute for the transfusion of blood, I have perused most of the literature on the subject, and find that many of the operations have been followed by marked success ; in some cases by temporary, and in others by permanent benefit ; but sometimes no good has accrued from the injection (when improperly performed, or when performed in unsuitable cases) ; and the operation appears to have proved fatal in a few instances. But, the intravenous injection of milk having now been practised[sic] under a considerable variety of circumstances, and many of the reports having been drawn up with great accuracy, it may perhaps prove of utility to lay before the profession an outline of the history of this important plan of treatment, which is now seldom resorted to, with the view of showing that the indications for this method of conveying nutrition are generally clear, and that, if properly performed, no harm can follow the operation.

- Charles Jennings, 1885

My current obsession is IV milk. We simply must talk about it.

The corpus of knowledge around this is a bit sparse. Joyously though, I bring you three articles: Pepper (1879), Meldon (1881), and Jennings (1885).

Despite Louis Pasteur rocking his stuff around 1860, it seems that pasteurization of milk was first suggested by Frans Von Soxhlet in 1886 (poor source), so it is likely that this milk was infused raw into these patients. In fact, discussion of this appears in Jennings (1885), which shows clear evidence for a marginal benefit of pasteurization - IV milk safety:

Dr. J. H. Howe experimented on seven dogs, substituting milk for the blood ; every dog died. He also tried the lacteal injection upon a man in the third stage of phthisis [pulmonary tuberculosis], in whom death occurred from coma soon afterwards. But Dr. Thomas, having performed the operation on the human subject with success, and having ascertained that the milk employed in Dr. Howe's experiments had been drawn from a cow one or one and a half hour's distance by rail from New York, reaching that city after a lapse of two or three hours, requested Dr. Eugene Dupuy to repeat the experiments upon dogs. Dr. Dupuy's experiments showed, 1, that the intravenous injection of decomposed milk into dogs is uniformly fatal ; 2, that the same experiment, if practised with perfectly pure and fresh milk, is entirely innocuous (ATciv York Medical Journal, May, 1878). The experiments of Professor Schafer also show that the intravenous injection of fresh milk in small quantities, or of milk boiled after standing, is harmless ; but that it is most dangerous to employ ordinary milk not so boiled, and the ordinary London milk is especially deleterious.

Sure, sometimes you infuse milk into someone and they get sepsis; also, sometimes you infuse it and the small particles get caught in capillary beds, properly messing up your pulmonary and renal circulation. Also, you may have an immune reaction and die. 

Speaking of immune reactions, Pepper details something that we would now call anaphylaxis in response to IV milk:

As soon as the milk began to enter the vein, there appeared violent capillary congestion of the face and surface of the body. The eyes were injected and prominent, the lips turgid, and the whole expression wild and alarming. The respiration was laboured, and an intense sense of oppression was felt, the patient clutching at the throat in her distress. While these symptoms were so severe, the funnel was lowered and the flow of milk arrested for a few seconds, when, as she became easier, it was resumed. Eighteen minutes after the operation, there was a sudden outbreak of urticaria, the wheals being large and pale-reddish... 

Luckily for him though, Jennings did not know much of that, and happily supports the use of IV milk. He describes the use of this procedure as a replacement for a blood transfusion (at this time, there was no understanding of blood typing). He was using milk as we would now use saline.

Meldon even tells us exactly how to go about about this whole ordeal:

...always add ten grains of carbonate of ammonia to each injection. This, besides rendering it certain that the milk is alkaline, prevents that depression which so frequently follows the operation. I invariably use goats' milk, as it is much more easy to bring that animal in close proximity to the patient, thus avoiding any unnecessary delay between milking and the injection. I never use more than six ounces of milk. When time will not allow fresh milk to be obtained, that at hand may be boiled and strained.

Milk as a replacement fluid

Let us detour into a modern approach to evaluating milk as an IV fluid.

Milk has an osmolality of 300mOsm/L, which makes it (like 0.9% saline) isotonic to blood. This is good, since it won’t lyse red cells.

In terms of its salts milk only has about 400mg of sodium per litre (which we can discount; one litre of milk would deliver the same amount of sodium as 50 millilitres of saline) but a whopping 1.4g of potassium (that’s 35mmol).

It also contains protein, which like human albumin will act via colloidal pressure to retain fluid in the intravascular space. How much protein? About 35g. How much volume expansion should we expect from that? Tricky to figure out but we can have a go. Let’s make a table (by stealing without apology from Dasta):

To work out the molar concentration of albumin, I just divided by the molar mass of human serum albumin which is 67kDa. For milk things are a lot trickier. There are multiple proteins in milk, mostly caesins, but on average the molar mass is close to 23kDa; SigmaAldrich are bizarrely helpful here. 

In terms of volume expansion, let’s consider infusing 1L of milk. Some of the osmoles of the milk are coming from potassium, which will rapidly move intracellularly and drag 35 milliosmoles worth of water with them (about 110mL of water at 300mOsm/L). There is basically no sodium in the solution to equilibrate across the capillary membrane, so the remaining ~265 milliosmoles are proteins and fat micelles. Assuming everything but the protein equilibrates across the capillary, we have introduced about twice as much onconic pressure as 5% albumin, so we should expect to induce about twice as much colloidal pressure, dragging fluid back into the capillary, to the tune of about 2L of intravascular volume per litre of milk. That means the six fluid ounce (170ml) advocated for by Meldon is the equivalent of about a litre of saline or Hartmann’s, which is nothing to sneeze at. Is milk more or less immunogenic than un-crossmatched human blood? Pretty hard to say. Neither is good for you.

So (I hear you murmur) perhaps the use of milk as an early IV fluid was not such a bad idea after all. And I might agree, except that actual proper IV crystalloids had existed since the cholera epidemic of 1831, some fifty years prior, where Latta pioneered it! He then promptly died of tuberculosis in 1833, and was along with his work so utterly forgotten that in the cholera return of 1850 hardly anyone received life-saving saline. It would be over a hundred years until his contribution was acknowledged – and decades before saline came around again.

Ring the bells that still can ring

Forget your perfect offering

There is a crack, a crack in everything

That’s how the milk gets in.

- Leonard Cohen (almost)

IV Milk as nutrition

Amazingly, Austin Meldon used IV milk as a treatment for pernicious anaemia (what we might, in our episteme, call vitamin B12 deficiency). In 1L of milk there’s 5microg of vitamin B12, which is admittedly only 0.5% of the intramuscular vitamin B12 dose we would give now, but might still allow one to squeeze out some new red blood cells for a time.

Modern day IV milk

All I can find of milk is a measly 30mL accidental injection of milk (compare the much more respectable 170ml recommended by Meldon). But joyously it appears we are still pumping people full of coconut water.

I wonder what the 2022 version of milk is, and which modern-day Latta we have forgotten.