Albumin is a globular protein with a molecular weight of 66-69,000 daltons (66-69 kDa). It is synthesized in the liver and is catabolized in various tissues where it is taken up by pinocytosis. Uptake and degradation is facilitated if the protein is structurally modified.  The constituent amino acids are utilized by the cells.


Albumin makes a large contribution to plasma colloid osmotic pressure due to its small size and abundance (35-50% of total plasma proteins by weight). It also serves as a carrier protein for many insoluble organic substances (e.g., unconjugated bilirubin, long chain fatty acids) and drugs. Being more anionic (negatively charged at physiologic pH), albumin also transports positively charged minerals, such as calcium (but does not serve as a body reservoir for calcium supplies), magnesium, zinc and copper. Aside from being a carrier for molecules, albumin is also thought to be an anti-oxidant protein by scavenging reactive oxygen species, protecting bound substances from oxidant injury, and binding free copper, which acts as an oxidant (Merlot et al 2014). Albumin contains no carbohydrate and is not stored to any significant extent by hepatocytes. The plasma half-life various among species (e.g. in humans it is 19 days [Merlot et al 2014]), and tends to increase with body size. Albumin can move into the extravascular space, either directly through gaps (e.g. sinusoidal endothelium in the liver) or across the endothelium with the help of receptors. Once in the extravascular space, albumin is taken up by lymphatics and removed back to the circulation. Small amounts are found in bodily fluids (eg. sweat) and urine (<1 mg/dL). Several receptors for albumin have been identified, some of which bind unmodified or native albumin or albumin altered by various natural and pathologic processes, such as glycosylation (e.g. diabetes mellitus) and oxidation. The latter processes enhance degradation of albumin, which likely occurs within macrophages via scavenger receptors (e.g. glycoprotein 18). The main receptor in continuous endothelium is albondin, which only binds native albumin, but scavenger receptors and other receptors also exist in endothelium and other tissues and these can bind native or altered albumin. The receptors in the proximal convoluted tubules in the kidney, where most of the albumin is resorbed after filtration, are the megalin-cubilin complex, which is deficient due to a genetic mutation in Border Collies and Beagles – the condition is called Imerslund-Gräsbeck syndrome and is characterized by defects in cobalamin intestinal absorption (where cubilin also has a major role) and proteinuria (Lutz et al 2013, Owczareck-Lipska et al 2013, Fyfe et al 2014).


Albumin is measured by its ability to bind to bromocresol green. Bromocresol purple is another dye that is used extensively in laboratories testing human samples, however this technique produces artifactually low values in animal sera and should not be used for measuring albumin in samples from animals. There are also species peculiarities in the amount of bromocresol green that binds to albumin. Falsely high albumin values are seen in samples from new world monkeys (e.g. lemurs) and rabbits, whereas low values are seen with birds. The falsely high values in rabbits can be overcome by the use of rabbit serum as a calibrator for the reaction (normally a human-based calibrator is used), however this is not routinely done by veterinary laboratories. Serum is the preferred sample for albumin measurement.

Reaction type

Blanked end-point


  • Bromocresol green (BCG) method: At an acidic pH of 4.1 albumin is significantly charged allowing it to bind to the anionic dye bromocresol green forming a blue-green complex. The color intensity of the complex is photometrically measured and is proportional to the concentration of albumin. Blanking indicates that a baseline reading is taken before the dye is added. This reduces falsely increased baseline absorbances, such as due to hemolysis or lipemia (as these are unchanged after addition of the dye and are subtracted from the results).
  • Reaction shown below:

albumin + bromocrescol green        pH 4.1      > blue-green complex

Units of measurement

Albumin concentration is measured in g/dL (conventional units) and g/L (SI units). The conversion equation is shown below:

g/dL x 10 = g/L

Sample considerations

Sample type

Serum and plasma (EDTA, heparin)

In one study, albumin concentrations were higher in serum than citrated plasma in sheep (Mohri and Rezapoor 2009). The mechanism in unclear and could be due to pH changes altering binding of the bromcresol green dye to the protein. Alternatively, it has been suggested that citrate, oxalate, and fluoride anticoagulants dilute protein concentration by causing water to diffuse from erythrocytes; therefore these anticoagulants should be avoided for samples collection. Liquid citrate will also dilute albumin by virtue of taking up 10% of blood volume. Lithium-heparin and K2-EDTA anticoagulants are preferred, however fibrinogen and other acute phase proteins may bind to bromcresol green resulting in higher concentrations in heparin than serum (Stokol et al 2001, Mohri and Rezapoor 2009).


The stability of albumin in human samples: 2.5 months at 15 – 25 °C, 5 months at 2 – 8 °C, and 4 months at (-15)-(-25) °C (per package insert, Roche).


  • Lipemia: Results may be affected with severely lipemic samples (lipemic index >550 units).
  • Hemolysis: Little effect on our method.
  • Icterus: Little effect on our method.
  • Drugs: Increases in albumin are reported in experimental studies in dogs administered corticosteroids. It is not clear if this is due to increased production of corticosteroids or dehydration secondary to free water losses from corticosteroid-induced polyuria. We have also made unpublished anecdotal associations between high albumin concentration and corticosteroid therapy in dogs.

Test interpretation

Changes in albumin concentration can selective (affecting albumin only) or non-selective (affecting both albumin and globulins). Selective versus non-selective changes in albumin and globulins yields potential clues as to the mechanism and cause for these abnormal results, as outlined below (the A:G ratio can be used as a guide as to selective versus non-selective changes).

Increased albumin concentration (hyperalbuminemia)

The most common cause of this is dehydration or volume contraction secondary to fluid loss.

  • Artifact: Albumin is higher in heparinized plasma than serum (due to non-specificity of bromocresol green which also binds to globulins, including fibrinogen [Stokol et al 2001]), however newer procedures have been optimized to minimize this phenomenon.
  • Physiologic: Hyperalbuminemia is a relative change seen with volume contraction secondary to fluid losses. Globulins may also increase in this situation, resulting in hyperproteinemia with no change in A:G ratio. However, globulins are frequently normal with volume contraction (low A:G), so the lack of hyperglobulinemia does not exclude dehydration or volume contraction as a cause of hyperalbuminemia. No changes in albumin concentration are seen up to 8 hours after feeding a regular meal to dogs (Yi et al 2022).
  • Iatrogenic: Prednisolone increased albumin concentrations by 15-17%, in a non-dose-dependent manner, in 8 Beagle dogs given 0.5, 1. 2 and 4 mg/kg/d for 5 days. It is possible this is due to corticosteroid-induced fluid losses through the kidney versus increased production in this short time frame (Tinklenberg et al 2020). We have noted associations between corticosteroid administration and high albumin (and transferrin) concentrations in dogs (unpublished observations). Albumin concentrations increased by a mean of 0.7-0.9 mg/dL in two groups of 7 cats given immunosuppressive doses of prednisolone (4.4 mg/kg/d) or dexamethasone (0.55 mg/kg/d) for 56 days (there was no untreated control group) (Lowe et al 2008). Mean urea nitrogen and creatinine concentrations also increased in both groups in the latter study, supporting a prerenal azotemia from fluid losses. In contrast, minimal changes in mean albumin concentrations were seen in 10 allergic cats given 1-2 mg/kg prednisolone for 13 days and the urine specific gravity was unchanged (Khelik et al 2019).
  • Pathophysiologic:
    • Adrenal dysfunction: In one study of 114 Scottish Terriers, high albumin concentrations (up to 5.2 g/dL) were observed in 22% of 61 dogs with clinical signs of hyperadrenocorticism and vacuolar hepatopathy (glycogen accumulation). Some dogs had a concurrent copper-associated hepatopathy. Hyperadrenocorticism was documented with a low dose dexamethasone test in around 1/4 of 46 tested dogs, with a high cortisol post ACTH stimulation in up to 2/3 of 37 tested dogs. Around 80% of 25 tested dogs had high adrenal sex hormones (progesterone and androstenedione) after ACTH stimulation (defined as a >1 fold increase above a reference interval).  Thus adrenal dysfunction, particularly affecting sex hormones, may be associated with a high albumin concentration, although the mechanism is unknown. Most of the dogs had high serum ALP activity (values up to 22, 000/uL) (Cortright et al 2014).
    • Hepatocellular carcinoma: A high albumin concentration (5.3 g/dL), with a concurrent increase in colloidal osmotic pressure, has been documented in a dog with hepatocellular carcinoma. Both these findings corrected after surgical removal of the tumor, although the mechanism for the tumor-associated increase in albumin was not ascertained (Cooper et al 2009). In the above study of 114 Scottish terriers with adrenal dysfunction and vacuolar hepatopathy, 10% of 28 tested dogs with concurrent hepatocellular carcinomas had high albumin (up to 4.5 g/dL), although not as high as those dogs without the tumor (see above). In this study, hematologic or serum biochemical findings did not differentiate between dogs with and without hepatocellular carcinomas (Cortright et al 2014).

Decreased albumin concentration (hypoalbuminemia)

  • Iatrogenic: Excessive fluid administration (overdilution).
  • Pathophysiologic
    • Decreased intake: Dietary deficiency of protein or anorexia could result in hypoalbuminemia from decreased protein intake with subsequent decreased albumin production. However, experimental studies of dietary protein deficiency in dogs indicates that this takes days to weeks (around 2 weeks) to result in albumin concentrations below the reference interval (Elman et al 1941). Due to continued production of globulins, dietary deficiency may only manifest with low protein concentrations 7-8 weeks after initiating the diet (Elman et al 1941). Decreased intake of protein is one of the contributing causes of hypoalbuminemia seen frequently in elderly humans (Brock et al 2016).
    • Decreased production:
      • Synthetic liver failure: Chronic hepatic disease will result in hypoalbuminemia when there is a > 80% reduction in functional mass. Other issues with protein production will frequently be apparent, e.g. low cholesterol concentration, low antithrombin activity, low protein C activity.
      • Acute phase reaction response: An acute phase reactant response is initiated in response to trauma, inflammation, neoplasia, etc and involves release of cytokines (IL-1, IL-6, TNFα) from macrophages. These cytokines act on regulatory elements in hepatocyte genes, resulting in upregulation of transcription of acute phase reactant proteins (fibrinogen, serum amyloid A, ceruloplasmin, haptoglobin) and downregulation of transcription of other proteins, including albumin and transferrin (so-called “negative acute phase reactants”). Increased degradation of albumin may also play a role in the hypoalbuminemia in this reaction, particularly if it is altered (e.g. by oxidation). In acute phase proteins, the A:G is decreased due to the combination of low albumin and high globulins (this type of response is associated with an increase in α2 globulins on serum electrophoresis). In one prospective study of 116 dogs admitted to an emergency center, dogs with systemic inflammation (systemic inflammatory response syndrome or sepsis) and dogs with localized inflammation had significantly lower albumin concentrations (1.9 ± 0.5 g/dL and 2.5 ± 0.4 g/dL, respectively) than clinically healthy blood donor dogs (3.2 ± 0.3 g/dL). Low albumin concentrations had good discriminatory ability between inflammation (localized or systemic) versus health (area under a receiver operator characteristic curve of 0.991-0.992) and between localized and systemic inflammation (area under the curve of 0.834), performing better than other markers of inflammation (low plasma iron, high fibrinogen, and high C-reactive protein concentrations) for the latter (Torrente et al 2015).
      • Increased oncotic pressure: The low albumin seen in some cases of multiple myeloma (systemic plasma cell tumor) has been attributed to increased oncotic pressure exerted by the immunoglobulins, particularly those causing hyperviscosity (IgM and high concentrations of some forms of IgG). The increased pressure is sensed by hepatic baroreceptors, which is then postulated to cause downregulation of liver production of albumin. However, other causes of low albumin (negative production from an acute phase response, glomerular protein loss) are operating with this tumor and may be primarily responsible for the low albumin concentration. 
    • Loss: The most common cause of low albumin and low globulins (non-selective protein loss) are protein-losing enteropathy and severe hemorrhage. Severe decreases in albumin alone, without concurrent change in globulins, should prompt a search for albumin loss through the kidney (glomerular disease) or decreased liver production (the latter is often accompanied by increased globulins, because the liver normally clears antigens derived from the gut and defective liver clearance of antigens can result in a polyclonal gammopathy). Mild to moderate decreases in albumin concentration are seen in horses with gastrointestinal disease (not protein-losing enteropathy per se, which is rare in this species) and may be the only abnormality on the chemistry panel.
      • Protein-losing enteropathies: In these conditions, albumin and globulins are often lost concurrently, thereby maintaining a normal A:G. There are exceptions to this, e.g. Basenjis with immunoproliferative bowel disease have hyperglobulinemia and some dogs with inflammatory bowel disease may not have concurrent decreases in both albumin and globulins. We often see hypocholesterolemia and low total iron binding capacity in protein-losing enteropathies. Hypocholesterolemia is a useful clue.
      • Protein-losing glomerulopathy: This can result in hypoalbuminemia when protein loss is severe.  Nephrotic syndrome is characterized by moderate to marked proteinuria, hypoalbuminemia, hypercholesterolemia and edema. In protein-losing nephropathy, including that due to nephrotic syndrome, albumin is lost, but globulin levels are usually maintained, resulting in a low A:G. Not all animals with nephrotic syndrome are edematous. As indicated above, dogs with defects in the cubilin-megalin complex can also lose albumin through proteinuria.
      • Hemorrhage: Albumin concentrations are decreased with both internal and external hemorrhage and attributed to replacement of blood volume with interstitial protein-poor fluid in animals that become hypovolemic with sufficient blood loss. Albumin concentration is more reliably decreased than total protein or globulin concentrations in experimental studies of anemia in dogs, where anemia was induced by withdrawing large amounts of blood via venipuncture (external blood loss) (Elman 1939, Elman et al 1944). Albumin concentrations are more frequently decreased with external versus internal hemorrhage, where the protein is not “lost” per se and can be readily reabsorbed or reused. Thus, protein concentrations may be normal, depending on the degree and duration of hemorrhage (internal, but also external). As a naturally occurring model of internal blood loss, retrospective studies of spontaneous hemoperitoneum in dogs and cats (e.g. from trauma or ruptured splenic hemangiosarcoma) revealed low protein concentrations in 42% and 36% of dogs (Lux et al 2013) and cats (Kulp et al 2010), respectively, with hypoalbuminemia being more common (78% of dogs and 55% of cats in the same studies). Thus, protein, albumin and globulin concentrations may not always be low with hemorrhage and depends on multiple factors, including the degree and duration of hemorrhage, need to maintain intravascular volume, the ability of the body to increase protein production in states of chronic or ongoing hemorrhage, and concurrent inflammation (which may depress albumin concentrations further, while increasing globulin concentrations. With severe hemorrhage (internal or external), albumin and globulins are expected to be concurrently low.
      • Severe exudative dermatopathies: This may also associated with concomitant albumin and globulin loss (A:G tends to remain normal), unless the dermatopathy stimulates an immune response (with hyperglobulinemia).
    • Sequestration: Hypoalbuminemia can be due to sequestration of albumin within body cavities in protein-rich effusions, e.g. peritonitis or inflammation resulting in increased vascular permeability. Globulins may be sequestered as well in these conditions. The low albumin concentration is also likely due to dilution from compensatory mechanisms associated with decreased effective circulating volume if fluid accumulation is severe, resulting in stimulation of ADH release and thirst. 
    • Catabolism: This is not a well-characterized mechanism for low albumin concentrations, although newer data support uptake of albumin in some cancer cells and altered albumin (e.g. oxidized) is more rapidly degraded than native or unaltered albumin (Merlot et al 2014). Increased albumin catabolism may occur with negative energy balance or protein malnutrition (e.g. chronic infections, neoplasia, trauma) and, potentially, as part of an acute phase response (see decreased production above). Older animals (and people) may also have hypoalbuminemia or hypoalbuminemia rapidly develops with poor nutrition or reduced food intake (Brock et al 2016). The mechanism is unclear, but may be related to decreased protein production as well as increased protein catabolism, potentially due to structural or chemical alterations in albumin that occur with age.
Scroll to Top