Interpretation of biochemical tests for iron deficiency: diagnostic difficulties related to limitations of individual tests

Most cases of iron deficiency can be diagnosed with simple tests. The concentration of serum iron does not fall until the body's iron stores are exhausted. As the stores are depleted, the concentration of transferrin rises while the concentration of ferritin falls. Caution is required when assessing patients with inflammatory disease as a low serum iron may not represent iron deficiency. These patients often have reduced concentrations of transferrin.

Biochemical tests for iron deficiency help to evaluate the cause of microcytic anaemia (a mean red cell corpuscular volume

Microcytic anaemia is most commonly due to iron deficiency, but is also caused by thalassaemia. It can also occur in some patients with anaemia secondary to chronic infection, inflammation, or malignancy (anaemia of chronic disease), even though the majority of these patients have a normal mean corpuscular volume (MCV).

The results of tests of iron status are relatively frequently distorted by other clinical factors. This is important to recognise as such distorted results may give a misleading view of the patient's iron stores. The impact of these factors can be recognised by combining the results of currently available tests. This gives a more reliable overview of the situation than is provided by any individual test. However, currently available tests provide a reliable index of iron status sufficiently frequently that it is not appropriate to perform endoscopic examinations merely because a patient has anaemia, especially normocytic anaemia. Endoscopy is appropriate only if there is evidence of iron deficiency not explained by other causes, or highly suggestive clinical indications of gastrointestinal disease.

Three tests provide most of the information the clinician needs:

– serum iron

– serum transferrin or iron binding capacity

– serum ferritin

Transferrin is the main iron transporting protein in the circulation. Ferritin concentrations reflect the body's iron stores.

The serum (or plasma) iron concentration falls progressively below the normal range (14-32 micromol/L) when the amount of iron in the body decreases after the reserves of iron have become exhausted. The concentration of transferrin rises under these circumstances towards, or above, the upper limit of the normal range. A subnormal level of iron in association with a supranormal level of transferrin is very strong evidence of iron deficiency (Table 1).

The diagnostic specificity of a low serum iron for iron deficiency is lost in the presence of inflammatory processes and certain other forms of chronic disease (Tables 2 and 3). The concentrations of iron and transferrin in the serum are significantly affected, and fall rapidly as part of the acute phase response after the onset of the inflammation irrespective of the status of the iron stores in the body. This response can produce a marked decrease within a day, especially in the serum iron level. The effect on iron and transferrin levels persists as long as the inflammatory process is sustained, and is classically associated with the development of anaemia of chronic disease. Unfortunately, a low serum iron level in this setting is frequently misinterpreted as evidence of iron deficiency, a major diagnostic error that can be avoided by simultaneous examination of the transferrin level, which in this context is subnormal or in the low normal range. Estimation of serum iron alone in the investigation of anaemia is consequently inadvisable.

Transferrin is the carrier protein which binds most of the iron present in serum. Its concentration in adults is estimated by immunological assays to be in the range of 2.0-3.6 g/L, with some minor variation depending on the particular method used. The total iron binding capacity of serum measured by determining the total amount of iron that can be bound to serum protein is proportional to the concentration of transferrin in the serum. Consequently, the total iron binding capacity simply represents an alternative means of expressing the amount of transferrin in serum.

The percentage occupation of the iron binding sites on transferrin by iron is calculated by dividing the serum iron level by the serum total iron binding capacity. The serum total iron binding capacity can be extrapolated from the transferrin level or measured directly. In normal individuals, the transferrin saturation is 20-50%. The view has been expressed that a subnormal percentage saturation is a useful index of iron deficiency, but low values are also obtained in chronic disorders, and consequently lack specificity.

Subnormal levels of ferritin can be detected when iron stores are exhausted, but before the serum iron level has become affected. Ferritin thus represents the most sensitive index of early iron deficiency. The normal range of ferritin in serum is dependent on several variables including methodology, age and sex. The normal range is 25-155 microgram/L in menstruating adult females, and 40-260 microgram/L in adult males. Concentrations in postmenopausal females correspond to those in males of a similar age, and are lower in children. A subnormal serum ferritin has a high degree of specificity for iron deficiency. However, in patients with iron deficiency, some clinical states can distort the relationship by causing the level of ferritin in the serum to increase (Table 4), although it is very uncommon for it to exceed 100 microgram/L in patients with iron deficiency.

Co-existing disease can sometimes make it impossible to assess iron status with biochemical tests. Concurrent measurement of serum iron, transferrin and ferritin reduces the chances of making an incorrect diagnosis of iron deficiency and, in most instances, provides a reliable index of body iron stores.

In the absence of unequivocal data on the status of iron reserves where such information is of major clinical importance, it is possible to assess iron reserves by performing a bone marrow aspirate. The marrow particles are stained for iron with the Prussian Blue reaction. A negative result is considered the gold standard for diagnosing iron deficiency, but the procedure is uncomfortable and is thus rarely performed.

Another approach is to assess the response to a trial of oral iron. A rise in the MCV, and an increase of >10 g/L in the haemoglobin level within 4-6 weeks is evidence that iron deficiency is contributing to the anaemia. Trials of iron supplements are, however, usually unrewarding in normocytic anaemia where an alternative mechanism is usually responsible for the anaemia.