This chapter is only peripherally related to Section I1(i) of the 2017 CICM Primary Syllabus, which asks the candidates to "explain the distribution and movement of body fluids". The movement of water in and out of cells seemed like an important part of this syllabus item, considering that this is where most of the body water is distributed. Relevant local links would probably also have to include the chapter which deals with the properties of the cell membrane (which the water would have to cross) and the chapter on the weird properties of intracellular fluid, which is where intracellular water resides (mainly as a hydrating gel adsorbed onto protein surfaces).
The reader is reminded that this item has never been much of a priority for the CICM examiners and will probably never appear in any of the exam papers. Reading more widely about this topic, therefore, would be an act of selfindulgence for the exam candidate, as indeed was the act of writing about it for the author. From this it follows that anybody who wants to read more would not be limited by the time constraints of exam preparation, and would be interested in a lot of detail from a recreational perspective, in the same spirit of vague curiosity with which one might watch a nature documentary. For these weirdos, the best reference would probably be Alan Finklestein's Water Movement Through Lipid Bilayers, Pores and Plasma Membranes: Theory and Reality (1987), a multi-volume monograph which is excellent, but which - as far as the author is able to establish - is only available on the shelves of large university libraries.
Water is not supposed to be lipid-soluble, and a lipid layer should be an absolute barrier to its passage. If not completely impermeable, the basic phospholipid palisade that makes up the cell membrane is supposed to be pretty tight: on its own, it probably has a permeability of around 1µm/s, according to some artificial membranes created and tested by Cass & Finkelstein (1967). Still, a surprisingly large amount of net water movement seems to occur between the intracellular and extracellular compartments. The main reason for this can be discerned from the most basic principles that govern diffusion. Consider:
Thus, even with very poor lipid solubility, a large amount of water is able to diffuse through the membrane, because there is so much water and there is so much membrane.
This series of totally theoretical observations above obviously refers only to some kind of idealised perfect bilayer of perfectly aligned phospholipid molecules. The cell membrane is of course nothing like this; it's probably more like a messy mosaic of large embedded molecules mortared together with phospholipid like a cobblestone pavement (Takamori et al, 2006). The water permeability of this surface will therefore vary from membrane to membrane, as cells vary in the number and properties of these surface molecules.
In fact even within the scale of the same single cell the membrane permeability properties may vary regionally, eg. between the apical and basolateral surfaces. Cells with lots of aquaporins in the membrane will have much higher water permeability; cells with especially cholesterol-rich membranes will have much lower permeability. According to some poorly referenced comments made by Reuss (2012), the range of permeabilities can span from almost zero (no permeability whatsoever) to 600 µm/s, which for a 7nm membrane represents essentially unimpeded water movement. The least permeable membranes are probably the thick ascending limb of the loop of Henle and the bladder epithelium, the latter made especially impermeable by rafts of uroplakin proteins in the membrane. The most permeable is probably in the cells of cortical collecting duct in the presence of abundant vasopressin.
Considering the freedom with which water seems to traffic in and out of these compartments, some equilibrium must exist, which keeps some water on one side of the membrane, and some on the other. This equilibrium is mainly determined by the osmolality of each compartment, and the movement of the water always trends towards achieving an equal osmolality on either side of the membrane. That means that the acute balance is mainly determined by the extracellular sodium concentration, because:
Extracellular sodium is therefore the main lever which you could pull to quickly adjust the balance of extracellular and intracellular fluid (the other being the retention or elimination of water, which indirectly adjusts extracellular sodium concentration). Intracellular osmolality and water content is also something that can be adjusted, and cells have various was of doing this (see the chapter on the mechanisms which maintain the intracellular fluid volume), but this process takes time.
Apart from sodium, there are of course other possible solutes. Clinically notable options include urea (which plays a role in the development of dialysis disequilibrium syndrome) and mannitol (which is administered with the specific intention of increasing extracellular osmolality and shrinking the intracellular fluid volume).