Medical Physiology, 3rd Edition

The Intracellular and Extracellular Fluids

Total-body water is the sum of the ICF and ECF volumes

Total-body water (TBW)imageN5-1 is ~60% of total-body weight in a young adult human male, ~50% of total-body weight in a young adult human female (Table 5-1), and 65% to 75% of total-body weight in an infant. TBW accounts for a lower percentage of weight in females because they typically have a higher ratio of adipose tissue to muscle, and fat cells have a lower water content than does muscle. Even if gender and age are taken into consideration, the fraction of total-body weight contributed by water is not constant for all individuals under all conditions. For example, variability in the amount of adipose tissue can influence the fraction. Because water represents such a large fraction of body weight, acute changes in TBW can be detected simply by monitoring body weight.


Approximate Water Distribution in Adult Humans*






Total-body water (TBW)

60% of BW


50% of BW


Intracellular fluid (ICF)

60% of TBW


60% of TBW


Extracellular fluid (ECF)

40% of TBW


40% of TBW


Interstitial fluid

75% of ECF


75% of ECF


Plasma (PV)

20% of ECF


20% of ECF


Transcellular fluid

5% of ECF


5% of ECF


Blood (BV)

PV/(1 − Hct)


PV/(1 − Hct)


*All of the above values are approximate and for illustration only. The volumes are rounded to the nearest liter, assuming a BW of 70 kg for both sexes, an Hct for men of 45%, and an Hct for women of 40%.

BW, body weight; Hct, hematocrit.


Determination of the Volume of Body Fluid Compartments

Contributed by Peter Aronson, Emile Boulpaep, Walter Boron

The TBW can be determined by the use of a volume-of-distribution technique. The first step is to infuse intravenously a known quantity of a tracer for water (deuterium oxide [2HOH] or tritiated water [3HOH]) that will distribute everywhere there is water. Because water readily permeates most cell membranes, a tracer for water distributes into both the extracellular and ICF.

For example, suppose a 70-kg male is injected with 108 counts per minute (cpm) of 3HOH contained in a small volume of physiological saline. After an equilibration period of 2 hours, a sample of the blood plasma is drawn and the 3HOH concentration in the plasma is found to be 2.5 × 103 cpm/mL plasma. Measurement also reveals that 5 × 105 cpm of 3HOH has been lost in urine, as well as from the skin and lungs. From this information, we can calculate the volume of distribution of the tracer, which is the same as the TBW:


(NE 5-1)

In this male, the TBW of 39.8 L is 57% of the 70-kg body weight.

How can we determine how this water is distributed among the various fluid compartments? In particular, we need to know the fraction of the TBW that is intracellular and the fraction that is extracellular. In practice, the ICF volume is calculated as the difference between TBW and ECF volume. The ECF volume is determined by using a marker that distributes uniformly throughout the compartments accessible to water but that does not enter the cells. Unfortunately, different markers thought to distribute within the ECF yield different values. For example, large polysaccharides (e.g., inulin) or polyalcohols (e.g., mannitol) that cannot cross cell membranes do not penetrate fully into dense connective tissue and bone. On the other hand, ions that are largely extracellular (e.g., Na+, Cl) are able to enter cells to some extent. Using the above techniques, the best estimate is that the total ECF represents between 20% and 25% of body weight (~40% of TBW). This leaves about 35% to 40% of the body weight (~60% of TBW) as intracellular water or ICF volume.

The plasma volume can be determined by measuring the volume of distribution of labeled albumin. Because albumin escapes only very slowly from the vascular compartment, measuring the final concentration of labeled albumin in the plasma and then using the above equation to compute the volume of distribution of the albumin label will yield the plasma volume.

The anatomy of the body fluid compartments is illustrated in Figure 5-1. The prototypical 70-kg male has ~42 L of TBW (60% of 70 kg). Of these 42 L, ~60% (25 L) is intracellular and ~40% (17 L) is extracellular. ECF is composed of blood plasma, interstitial fluid, and transcellular fluid.


FIGURE 5-1 Fluid compartments of a prototypical adult human male weighing 70 kg. TBW is divided into four major compartments: ICF (green), interstitial fluid (blue), blood plasma (red), and transcellular water such as synovial fluid (tan). Color codes for each of these compartments are maintained throughout this book.

Plasma Volume

Of the ~17 L of ECF, only ~20% (~3 L) is contained within the cardiac chambers and blood vessels, that is, within the intravascular compartment. The total volume of this intravascular compartment is the blood volume, ~6 L. The extracellular 3 L of the blood volume is the plasma volume. The balance, ~3 L, consists of the cellular elements of blood: erythrocytes, leukocytes, and platelets. The fraction of blood volume that is occupied by these cells is called the hematocrit. The hematocrit is determined by centrifuging blood that is treated with an anticoagulant and measuring the fraction of the total volume that is occupied by the packed cells.

Interstitial Fluid

About 75% (~13 L) of the ECF is outside the intravascular compartment, where it bathes the nonblood cells of the body. Within this interstitial fluid are two smaller compartments that communicate only slowly with the bulk of the interstitial fluid: dense connective tissue, such as cartilage and tendons, and bone matrix.

The barriers that separate the intravascular and interstitial compartments are the walls of capillaries. Water and solutes can move between the interstitium and blood plasma by crossing capillary walls and between the interstitium and cytoplasm by crossing cell membranes.

Transcellular Fluid

Finally, ~5% (~1 L) of ECF is trapped within spaces that are completely surrounded by epithelial cells. This transcellular fluid includes the synovial fluid within joints and the cerebrospinal fluid surrounding the brain and spinal cord. Transcellular fluid does not include fluids that are, strictly speaking, outside the body, such as the contents of the gastrointestinal tract or urinary bladder.

ICF is rich in K+, whereas ECF is rich in Na+ and Cl

Not only do the various body fluid compartments have very different volumes, they also have radically different compositions, as summarized in Figure 5-1Table 5-2 is a more comprehensive listing of these values. ICF is high in K+ and low in Na+ and Cl; ECF (interstitial and plasma) are high in Na+ and Cl and low in K+. The cell maintains a relatively high K+ concentration ([K+]i) and low Na+ concentration ([Na+]i), not by making its membrane totally impermeable to these ions but by using the Na-K pump to extrude Na+ actively from the cell and to transport K+ actively into the cell.


Approximate Solute Composition of Key Fluid Compartments






Na+ (mM)





K+ (mM)





Ca2+ (mM)

1.2 (ionized)
2.4 (total)*

1.3 (ionized)

1.2 (ionized)

0.0001 (ionized)

Mg2+ (mM)

0.6 (ionized)
0.9 (total)*

0.6 (ionized)

0.55 (ionized)

1 (ionized)
18 (total)

Cl (mM)





image (mM)





image and image (mM)

0.7 (ionized)
1.4 (total)

0.75 (ionized)

0.8 (ionized)

0.7 (free)


7 g/dL
1 mmole/L
14 meq/L

1 g/dL

30 g/dL

Glucose (mM)




Very low






Osmolality (milliosmoles/kg H2O)





*Total includes amounts ionized, complexed to small solutes, and protein bound.

Arterial value. The value in mixed-venous blood would be ~24 mM.

As discussed on pages 1054–1056, levels of total plasma inorganic phosphate are not tightly regulated and vary between 0.8 and 1.5 mM.

Transcellular fluids differ greatly in composition, both from each other and from plasma, because they are secreted by different epithelia. The two major constituents of ECF, the plasma and the interstitial fluid, have similar composition as far as small solutes are concerned. For most cells, it is the composition of the interstitial fluid enveloping the cells that is the relevant parameter. The major difference between plasma and interstitial fluid is the absence of plasma proteins from the interstitium. These plasma proteins, which cannot equilibrate across the walls of most capillaries, are responsible for the usually slight difference in small-solute concentrations between plasma and interstitial fluid. Plasma proteins affect solute distribution because of the volume they occupy and the electrical charge they carry.

Volume Occupied by Plasma Proteins

The proteins and, to a much lesser extent, the lipids in plasma ordinarily occupy ~7% of the total plasma volume. Clinical laboratories report the plasma composition of ions (e.g., Na+, K+) in units of milliequivalents (meq) per liter of plasma solution. However, for cells bathed by interstitial fluid, a more meaningful unit would be milliequivalents per liter of protein-free plasma solution because it is only the protein-free portion of plasma—and not the proteins dissolved in this water—that can equilibrate across the capillary wall. For example, we can obtain [Na+] in protein-free plasma (which clinicians call plasma water) by dividing the laboratory value for plasma [Na+] by the plasma water content (usually 93%):



Similarly, for Cl,



Table 5-2 lists solute concentrations in terms of both liters of plasma and liters of plasma water. If the plasma water fraction is <93% because of hyperproteinemia (high levels of protein in blood) or hyperlipemia (high levels of lipid in blood), the values that the clinical laboratory reports for electrolytes may appear abnormal even though the physiologically important concentration (solute concentration per liter of plasma water) is normal. For example, if a patient's plasma proteins and lipids occupy 20% of plasma volume and consequently plasma water is only 80% of plasma, a correction factor of 0.80 (rather than 0.93) should be used in Equation 5-1. If the clinical laboratory were to report a very low plasma [Na+] of 122 meq/L plasma, the patient's [Na+] relevant to interstitial fluid would be 122/0.80 = 153 meq/L plasma water, which is quite normal.

Effect of Protein Charge

For noncharged solutes such as glucose, the correction for protein and lipid volume is the only correction needed to predict interstitial concentrations from plasma concentrations. Because plasma proteins carry a net negative charge and because the capillary wall confines them to the plasma, they tend to retain cations in plasma. Thus, the cation concentration of the protein-free solution of the interstitium is lower by ~5%. Conversely, because these negatively charged plasma proteins repel anions, the anion concentration of the protein-free solution of the interstitium is higher by ~5%. We consider the basis for these 5% corrections in the discussion of the Gibbs-Donnan equilibrium (see pp. 128–129).

Thus, for a monovalent cation such as Na+, the interstitial concentration is 95% of the [Na+] of the protein-free plasma water, the value from Equation 5-1:



For a monovalent anion such as Cl, the interstitial concentration is 105% of the [Cl] of the protein-free water of plasma, a value already obtained in Equation 5-2:



Thus, for cations (e.g., Na+), the two corrections (0.95/0.93) nearly cancel each other. On the other hand, for anions (e.g., Cl), the two corrections (1.05/0.93) are cumulative and yield a total correction of ~13%.

All body fluids have approximately the same osmolality, and each fluid has equal numbers of positive and negative charges


Despite the differences in solute composition among the intracellular, interstitial, and plasma compartments, they all have approximately the same osmolality. Osmolality describes the total concentration of all particles that are free in a solution. imageN5-2 Thus, glucose contributes one particle, whereas fully dissociated NaCl contributes two. Particles bound to macromolecules do not contribute at all to osmolality. In all body fluid compartments, humans have an osmolality—expressed as the number of osmotically active particles per kilogram of water—of ~290 milliosmoles/kg H2O (290 mOsm).


Osmolality versus Osmolarity

Contributed by Peter Aronson, Emile Boulpaep, Walter Boron

Osmolality is a measure of the number of osmotically active particles per kilogram of H2O. The number of particles is expressed in units of moles. Thus, 1 osmole is 1 mole of osmotically active particles. Note that we express osmolality in terms of the mass of solvent (H2O), not the mass of the entire solution (i.e., solutes and solvent). Unfortunately, it is rather impractical to measure the mass of H2O in a solution (e.g., you could weigh the material before and after evaporating all the H2O). For that reason, chemists have introduced osmolarity, the number of osmotically active particles per liter of total solution. It is easy to determine this volume. For very dilute solutions, the osmolality and osmolarity are quantitatively almost identical. Even for interstitial fluid, osmolality and osmolarity differ by <1%. Thus, for all practical purposes, one could use these terms interchangeably. On the other hand, the osmometers used to determine the number of osmoles in body fluids are usually calibrated with standards that are labeled in terms of osmoles per kilogram of H2O (i.e., osmolality). Therefore, in this text, we express the osmotic activity of solutions in terms of osmolality.

Blood plasma presents a special problem. Plasma proteins occupy ~7% of the total volume of plasma, but cannot cross the capillary wall. The solution that equilibrates across the capillary wall is the protein-free part of the blood plasma, which clinicians refer to as “plasma H2O.” Therefore, the osmolality of the interstitial fluid will be the same as the osmolality of the protein-free portion of blood plasma. This value is ~290 milliosmoles/kg or 290 mOsm. The osmolality of the total volume of the blood plasma (i.e., the protein-free portion plus the proteins) is only 291 mOsm. The extra 1 mOsm is the osmotic pressure of the plasma proteins. This extra 1 mOsm has a special name: colloid osmotic pressure or oncotic pressure (p. 128). The reason that the plasma proteins contribute so little is that—although they have a large mass—they have a high molecular weight and thus represent very few particles.

Plasma proteins contribute ~14 meq/L (see Table 5-2). However, because these proteins usually have many negative charges per molecule, not many particles (~1 mM) are necessary to account for these milliequivalents. Moreover, even though the protein concentration—measured in terms of grams per liter—may be high, the high molecular weight of the average protein means that the protein concentration—measured in terms of moles per liter—is very low. Thus, proteins actually contribute only slightly to the total number of osmotically active particles (~1 mOsm).

Summing the total concentrations of all the solutes in the cells and interstitial fluid (including metabolites not listed in Table 5-2), we would see that the total solute concentration of the intracellular compartment is higher than that of the interstitium. Because the flow of water across cell membranes is governed by differences in osmolality across the membrane, and because the net flow is normally zero, intracellular and extracellular osmolality must be the same. How, then, do we make sense of this discrepancy? For some ions, a considerable fraction of their total intracellular store is bound to cellular proteins or complexed to other small solutes. In addition, some of the proteins are themselves attached to other materials that are out of solution. In computing osmolality, we count each particle once, whether it is a free ion, a complex of two ions, or several ions bound to a protein. For example, most of the intracellular Mg2+ and phosphate and virtually all the Ca2+ are either complexed or bound. Some of the electrolytes in blood plasma are also bound to plasma proteins; however, the bound fraction is generally much lower than the fraction in the cytosol.


All solutions must respect the principle of bulk electroneutrality: the number of positive charges in the overall solution must be the same as the number of negative charges. If we add up the major cations and anions in the cytosol (see Table 5-2), we see that the sum of [Na+]i and [K+]i greatly exceeds the sum of [Cl]i and image. The excess positive charge reflected by this difference is balanced by the negative charge on intracellular macromolecules (e.g., proteins) as well as smaller anions such as organic phosphates.

There is a similar difference between major cations and anions in blood plasma, where it is often referred to as the anion gap. The clinical definition of anion gap is



Note that plasma [K+] is ignored. The anion gap, usually 9 to 14 meq/L, is the difference between ignored anions and ignored cations. Among the ignored anions are anionic proteins as well as small anionic metabolites. Levels of anionic metabolites, such as acetoacetate and β-hydroxybutyrate, can become extremely high, for example, in type 1 diabetic patients with very low levels of insulin (see Box 51-5). Thus, the anion gap increases under these conditions.

The differences in ionic composition between the ICF and ECF compartments are extremely important for normal functioning of the body. For example, because the K+ gradient across cell membranes is a major determinant of electrical excitability, clinical disorders of extracellular [K+] can cause life-threatening disturbances in the heart rhythm. Disorders of extracellular [Na+] cause abnormal extracellular osmolality, with water being shifted into or out of brain cells; if uncorrected, such disorders lead to seizures, coma, or death.

These examples of clinical disorders emphasize the absolute necessity of understanding the processes that control the volume and composition of the body fluid compartments. These processes are the ones that move water and solutes between the compartments and between the body and the outside world.