Iron deficiency is the state in which the content of iron in the body is less than normal. Iron depletion is the earliest stage of iron deficiency, in which storage iron is decreased or absent but serum iron concentration, transferrin saturation, and blood hemoglobin levels are normal. Iron deficiency without anemia is a somewhat more advanced stage of iron deficiency, characterized by absent storage iron, usually low serum iron concentration and transferrin saturation, but without frank anemia. Iron-deficiency anemia, the most advanced stage of iron deficiency, is characterized by absent iron stores, low serum iron concentration, low transferrin saturation, and low blood hemoglobin concentration.
Chlorosis, or “green sickness,” was well known to European physicians after the middle of the 16th century. In France, by the middle of the 17th century, iron salts and other remedies (including, oddly enough, phlebotomy) were used in its treatment. Not long thereafter, iron was recommended by Sydenham as a specific remedy for chlorosis. For the 100 years preceding 1930, iron was used in the treatment of chlorosis, often in ineffective doses, although the mechanism of action of iron and the appropriateness of its use were highly controversial. By the beginning of the 20th century, it had been established that chlorosis was characterized by a decrease in the iron content of the blood and by the presence of hypochromic erythrocytes, but it was not until the classic 1932 studies by Heath, Strauss, and Castle1 that it was shown that the response of anemia to iron was stoichiometrically related to the amount of iron given and that chlorosis was, indeed, iron deficiency. The history of iron deficiency has been reviewed in greater detail elsewhere.2,3
Iron-deficiency anemia is the most common anemia worldwide, and is especially prevalent in women and children in regions where meat intake is low, food is not fortified with iron, and malaria, intestinal infections, and parasitic worms are common.4,5,6 Women with frequent pregnancies may be particularly susceptible. In the United States, iron deficiency is most common in children 1 to 4 years old and in adolescent, reproductive age, or pregnant women.7,8,9
ETIOLOGY AND PATHOGENESIS
Iron deficiency may occur as a result of chronic blood loss, diversion of iron to fetal and infant erythropoiesis during pregnancy and lactation, inadequate dietary iron intake, malabsorption of iron, intravascular hemolysis with hemoglobinuria, diversion of iron to nonhematopoietic tissues like the lung, genetic factors, or a combination of these factors. Of these, gastrointestinal or menstrual blood loss are the most common. As discussed in Chap. 42, the average adult male has approximately 1000 mg of iron in stores, but on average, women have less than half of this amount. The average daily dietary intake of iron is 10 to 12 mg, but much of this is not absorbed, even when absorption is maximal. Blood loss of each milliliter of packed erythrocytes represents 1 mg of iron. Thus chronic daily blood loss greater than 5 mL of erythrocytes will deplete iron reserves over weeks to months, and even if bleeding stops completely, the repletion of lost iron, including the restoration of iron stores (around 1000 mg in the average adult man), will take many months.
Gastrointestinal Blood Loss
In men and in postmenopausal women, iron deficiency is most commonly caused by chronic bleeding from the gastrointestinal tract. Table 43–1 lists the causes of such blood loss. After history and physical examination rule out an obvious bleeding source in the genitourinary or respiratory tracts, evaluation of the gastrointestinal tract10 is necessary because of the potential that the pathologic process causing the blood loss is life-threatening. In the adult, the most common causes are peptic ulcer, erosion in a hiatal hernia, gastritis (including that caused by alcohol or aspirin ingestion), hemorrhoids, vascular anomalies (such as angiodysplasia), and neoplasms.
Table 43–1.Sources of Blood Loss ||Download (.pdf) Table 43–1. Sources of Blood Loss
|ALIMENTARY TRACT |
|Stomach and duodenum |
Leiomyoma (Ménétrier disease)
Antral vascular ectasia
|Small intestine |
|Colon and anorectal |
|BILIARY TRACT |
|GENITOURINARY TRACT |
|RESPIRATORY TRACT |
Idiopathic pulmonary hemosiderosis
Gastritis, Varices, Ulcers, and Inflammation
Gastritis as a result of drug ingestion is a common cause of bleeding. Aspirin, indomethacin, ibuprofen, and other nonsteroidal antiinflammatory drugs cause gastritis, but may also cause bleeding by inducing gastric or duodenal ulcers, or lesions in the small intestine11 and even the colon. Gastritis caused by alcohol ingestion12 can also cause significant blood loss. Chronic blood loss is often the cause of anemia in rheumatoid arthritis (perhaps because of the use of nonsteroidal antiinflammatory medications) and in inflammatory bowel disease.
Chronic blood loss from esophageal or gastric varices can lead to iron-deficiency anemia. Hemorrhoidal bleeding may lead to severe iron-deficiency anemia. Chronic blood loss may result from diffuse gastric mucosal hypertrophy (Ménétrier disease).13 Peptic ulcers of the stomach or duodenum are common causes of iron deficiency, and an association between infection with Helicobacter pylori and iron-deficiency anemia has been documented in numerous studies.14 Some of these iron-deficient patients who are infected with H. pylori do not respond to oral iron alone, but do respond to eradication of H. pylori.15
Gastric ulceration and bleeding can also occur in disorders of hypergastrinemia, as in Zollinger-Ellison syndrome and pseudo–Zollinger-Ellison syndrome. Although concerns were raised that long-term medical therapy of these disorders with proton pump inhibitors would also cause iron deficiency by raising gastric pH and making iron less soluble, this does not seem to be the case.16 Anemia that follows subtotal gastrectomy is usually attributed to reduced absorption of dietary iron (see “Malabsorption of Iron” below), but occult intermittent gastrointestinal bleeding from gastrointestinal lesions may also be a contributory factor, and requires endoscopic evaluation.17
Diaphragmatic (hiatal) hernia is often associated with gastrointestinal bleeding.18,19,20 The frequency of anemia ranges from 8 to 38 percent. Bleeding is much more likely to occur in patients with paraesophageal or large hernias than in those with sliding or small hernias. Mucosal changes cannot always be demonstrated by esophagoscopy or gastroscopy in patients who have had blood loss from hiatus hernia. However, a linear gastric erosion, also called a “Cameron ulcer,” commonly occurs on the crests of mucosal folds at the level of the diaphragm, and appears to be the site of bleeding.21
Hookworms are a major cause of gastrointestinal blood loss in many parts of the world.22
The lesions of angiodysplasia may occur in any part of the gastrointestinal tract.23 These tiny vascular anomalies may be the cause of significant blood loss. Endoscopy is usually required for diagnosis, and often needs to be repeated as bleeding can be intermittent. Gastric antral vascular ectasia24 exhibits a characteristic endoscopic appearance (“watermelon stomach”), and is another cause of blood loss. Hemorrhage into the biliary tract is a rare cause of chronic iron-deficiency anemia.25
Tortuous, dilated sublingual venous structures, the cherry hemangiomas commonly seen in the elderly, and the spider telangiectases of chronic liver disease are usually easily distinguished from the lesions of hereditary hemorrhagic telangiectasia. Bleeding from intestinal telangiectases has also been observed in scleroderma26 and in Turner syndrome,27 as a manifestation of bleeding from abnormal blood vessels. Cutaneous hemangiomas (blue rubber bleb nevus) may be associated with hemorrhage from intestinal hemangiomas.28
In hereditary hemorrhagic telangiectasia (Chap. 122), characteristic lesions commonly occur on fingertips, nasal septum, tongue, lips, margins (helices) of ears, oral and pharyngeal mucosa, palms and soles, and other epithelial and cutaneous surfaces throughout the body. Those lesions that occur in the gastrointestinal tract are particularly likely to bleed and to cause iron deficiency.
Meckel diverticulum is a very common abnormality representing a vestigial remnant of the omphalomesenteric duct. In children, bleeding from this structure accounts for a small proportion of cases of iron-deficiency anemia.29
Heavy menstrual bleeding30 is a very common cause of iron deficiency. The amount of blood lost with menstruation31 varies markedly from one woman to another and is often difficult to evaluate by questioning the patient. The average menstrual blood loss is approximately 40 mL per cycle. Blood loss exceeds 80 mL (equivalent to approximately 30 mg of iron) per cycle in only 10 percent of women. The volume of blood lost in the course of one menstrual cycle may be as high as 495 mL in apparently healthy, nonanemic women who do not regard their menstrual flow to be excessive. The amount of menstrual blood lost does not seem to vary markedly from one cycle to another for any given individual. Oral contraceptives reduce menstrual blood loss, but the use of an intrauterine coil for contraception increases menstrual blood loss, especially during the first year of use. Because the daily dietary intake is usually between 10 and 12 mg of iron and only a few milligrams of this can be absorbed, iron balance in many menstruating women is precarious.
Excessive bleeding may be caused by uterine fibroids and malignant neoplasms. Neoplasms, stones, or inflammatory disease of the kidney, ureter, or bladder may cause enough chronic blood loss to produce iron deficiency.
In the absence of hematuria, urinary iron losses as high as 1 mg/day have been reported in rare patients with nephrotic syndrome, some of whom had hypoferremia and hypochromic anemia.32 We found only one report of a patient in whom abnormally high urinary iron loss may have caused anemia without proteinuria or hematuria.33
Hemostatic defects, particularly those related to abnormal platelet function or number may lead to gastrointestinal bleeding, although unless the thrombocytopenia or platelet dysfunction is severe, gastrointestinal bleeding usually signifies an abnormality in the gastrointestinal tract. Gastrointestinal bleeding is common in von Willebrand disease (Chap. 126), but often because of coexistent peptic ulcer disease. Polycythemia vera is typically associated with iron deficiency as a result either of spontaneous gastrointestinal hemorrhage that commonly occurs in this disorder, or phlebotomy therapy, or both (Chap. 84).
When a patient with a disorder of hemostasis suffers from gastrointestinal bleeding, one must consider the possibility that the bleeding may not be caused by a hemostatic defect alone, but that an anatomic lesion of the gastrointestinal tract may also be present.
Nosocomial (Iatrogenic) Anemia
Iatrogenic anemia is particularly prevalent in intensive care units34 where repetitive blood sampling may result in removal of 40 to 70 mL of blood daily, and this iatrogenic phlebotomy can result in iron-deficiency anemia.
The use of extracorporeal dialysis for treatment of chronic renal disease may cause iron deficiency,35,36 often superimposed upon the anemia of chronic renal disease (Chap. 37). Patients treated with chronic hemodialysis experience multiple sources of blood loss with the dialysis equipment is a major cause, along with gastrointestinal bleeding, blood sampling and bleeding related to vascular access.
Anemia Incident to Blood Donation
Each whole-blood donation removes approximately 200 mg of iron from the body. Lesser amounts of iron are removed in the course of donating platelets or leukocytes. Potential donors are screened in blood banks, so that those with frank anemia are not phlebotomized. Yet, by the time they are excluded from donation, some blood donors are iron depleted37,38,39 and may readily develop iron-deficiency anemia with relatively small additional blood loss.
Factitious anemia as a result of self-inflicted bleeding may present a formidable diagnostic and therapeutic problem. This rare condition has also been called, in literary allusion to a fictitious character, “Lasthénie de Ferjol syndrome” (in Barbey d’Aurevilly’s gloomy novel, Une Histoire Sans Nom), or part of Munchausen syndrome (based on the Rudolf Raspe book, The Surprising Adventures of Baron Münchausen).40,41 Most patients are women, and are often employed in a medical setting. There is often a history of numerous blood transfusions. The anemia is chronic and may be severe. The site of induced blood loss is obscure. Hence, patients are subjected to numerous radiographic and endoscopic examinations, usually to no avail. The patients are usually refractory to medical advice and therapy. The patients may be depressed and suicidal; some also suffer anorexia nervosa. Psychiatric care is needed, but often is unsuccessful. Rarely, the outcome of self-bleeding may be fatal.42
Ingestion of whole cow’s milk may induce protein-losing enteropathy and gastrointestinal bleeding in infants,43,44 probably on the basis of hypersensitivity or allergy. In four such cases observed endoscopically, erosive gastritis or gastroduodenitis was demonstrated as the probable source of bleeding. At least during the first year of life, children should not be given whole bovine milk, either raw or pasteurized. More protracted heating, as in preparation of infant formulas, eliminates this problem. Intrinsic lesions of the gastrointestinal tract, such as those listed above, may cause bleeding in infants, as well as in older children.
Persistent recurrent hemoptysis may lead to iron-deficiency anemia. It may be a result of congenital anomalies of the respiratory tract, endobronchial vascular anomalies, chronic infections, neoplasms, or valvular heart disease. Severe iron-deficiency anemia is a manifestation of idiopathic pulmonary hemosiderosis45 and of Goodpasture syndrome (progressive glomerulonephritis with intrapulmonary hemorrhage). In some of these disorders, hemoptysis may not be observed, but sufficient amounts of blood-laden sputum may be swallowed to result in positive tests for occult blood in the stools. Iron deficiency occurs in a large proportion of patients with cystic fibrosis,46,47 and occurs even in the absence of hemoptysis, suggesting that inflammatory inhibition of dietary iron absorption and iron loss in purulent sputum could contribute to the deficiency.
Pregnancy and Parturition
Although physiologic decrease in hemoglobin concentration is an expected consequence of hemodilution associated with pregnancy, true iron deficiency frequently results in more severe anemia. In pregnancy, the average iron loss resulting from diversion of iron to the fetus, blood loss at delivery (equivalent to an average of 150 to 200 mg of iron), and lactation is altogether approximately 900 mg; in terms of iron content, this is equivalent to the loss of more than 2 L of blood. Approximately 30 mg of iron may be expended monthly in lactation. Because most women begin pregnancy with low iron reserves, these additional demands frequently result in iron-deficiency anemia. Iron depletion has been reported in some 85 to 100 percent of pregnant women. Iron-deficient mothers are likely to have smaller babies. The incidence of anemia and iron deficiency is lower in women who take oral iron supplementation, daily or intermittently.48,49,50,51 In regions with endemic malaria, iron supplementation may increase the risk of malaria and some recommend that it be combined with malarial prophylaxis.52 Most experts agree that oral iron supplementation during pregnancy is desirable despite side effects. Increasing safety and convenience of parenteral iron therapy may lead to reevaluation of its role in the prevention and treatment of iron-deficiency anemia of pregnancy.53
In infants, iron deficiency is most often a result of the use of unsupplemented milk diets, which contain an inadequate amount of iron. During the first year of life, the full-term infant requires approximately 160 mg and the premature infant approximately 240 mg of iron to meet the needs of an expanding red cell mass. Approximately 50 mg of this need is fulfilled by the destruction of erythrocytes that occurs physiologically during the first week of life (Chaps. 7 and 33); the rest must come from the diet. Milk products are very poor sources of iron, and prolonged breast- or bottle-feeding of infants frequently leads to iron-deficiency anemia unless iron supplementation is implemented. This is especially true of premature infants. The European Society for Pediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN) Committee on Nutrition urges that all infant formulas be iron-fortified54; in North America, the use of iron-fortified formula is now generally accepted, but there is controversy about the appropriate level of fortification.55 In older children, an iron-poor diet may also contribute to the development of iron-deficiency anemia, particularly during rapid growth periods.
Infants and young women are usually in precarious iron balance, their iron intake being less than 80 percent of the recommended daily allowance (RDA).56 Fortification of bread and cereals with ferrous sulfate or metallic iron is commonplace. This practice was suspended in Sweden because of concern for the possibility of increasing iron storage in patients with the hemochromatosis genotype, resulting in increased incidence of iron-deficiency anemia.57
The scant iron supply of the American diet places young women and children at particular risk of negative iron balance. Because the adult male needs to absorb only approximately 1 mg iron daily from his diet to maintain normal iron balance, iron deficiency in older men is very rarely caused by insufficient dietary intake alone.
Gastric secretion of hydrochloric acid is often reduced in iron deficiency.58 Histamine-fast achlorhydria has been found in as many as 43 percent of patients with iron deficiency. Gastric function may improve after correction of the iron deficiency, so that iron deficiency may be both a cause and a result of impairment of gastric iron secretion. However, in persons older than the age of 30 years, the achlorhydria is usually irreversible. Furthermore, when atrophic gastritis coexists with iron deficiency, no improvement in gastric secretory function has followed iron therapy. Autoimmune gastritis, which is often associated with H. pylori infection,14,15 may play an important role in both iron-deficiency anemia and, in later life, in the development of pernicious anemia.
Intestinal malabsorption of iron is quite an uncommon cause of iron deficiency except after gastrointestinal surgery and in malabsorption syndromes. Ten to 34 percent of patients who have undergone subtotal gastric resection develop iron-deficiency anemia years later. Many such patients have impaired absorption of food iron, caused in part by more rapid gastrojejunal transit and in part by partially digested food bypassing some of the duodenum as a result of the location of the anastomosis. Fortunately, medicinal iron is well absorbed in post–partial gastrectomy patients. Moreover, gastrointestinal blood loss may also play an important role in anemia following gastric resection (see “Gastrointestinal Blood Loss” earlier). In malabsorption syndromes, absorption of iron may be so limited that iron-deficiency anemia develops over a period of years. Celiac disease, whether overt or occult, may be associated with iron-deficiency anemia.14,15,59,60
Intravascular Hemolysis and Hemoglobinuria
Iron-deficiency anemia may occur in paroxysmal nocturnal hemoglobinuria (Chap. 40) and in hemolysis resulting from mechanical erythrocyte trauma from intracardiac myxomas, valvular prostheses (particularly if malfunctioning), or patches (Chaps. 33 and 51). In these disorders, up to 10 mg/day of iron is lost in the urine as hemosiderin and ferritin in desquamated tubular cells and as hemoglobin dimers, an amount sufficient to cause systemic iron deficiency.61,62
Iron deficiency occurs frequently in athletes engaged in a variety of sports (Chaps. 33 and 51), especially female athletes.63 There may be mild anemia. Increased intravascular hemolysis, presumably with some renal loss of iron, may play a role, but gastrointestinal blood loss has also been demonstrated in persons engaged in strenuous athletic pursuits. Hemoglobinuria and hemosiderinuria are also seen in competitive and recreational runners, that is, march hemoglobinuria (Chaps. 33 and 51). Strenuous exercise also elicits a rise in serum interleukin (IL)-6 and hepcidin, and this could decrease dietary iron absorption.63
Women soldiers undergoing basic training experience iron depletion as determined by serum ferritin measurements, and this can be partially reversed by iron supplementation.64 The etiology may be similar to the iron deficiency seen in athletes.
Based on twin studies,65 genetic factors play a role in iron deficiency. Mutations in multiple genes, including HFE and transferrin, show weak associations with iron stores but only mutations of the membrane serine protease Tmprss666 have been identified in genome-wide association studies as genetic factors that cause or predispose to iron deficiency. The genetic syndrome of iron-refractory iron deficiency anemia is mediated by inappropriately increased hepcidin as a result of homozygous or compound heterozygous mutations in Tmprss6.67,68,69 Increased hepcidin diminishes iron absorption and causes inappropriate retention of available iron in splenic macrophages and Kupffer cells.
As iron deficiency develops, different compartments are depleted in iron in an overlapping sequence, as illustrated schematically in Fig. 43–1.
Stages in the development of iron deficiency. Early iron deficiency (iron depletion) is usually not accompanied by any abnormalities in blood; at this stage, serum iron concentration is occasionally below normal values and storage iron is markedly depleted. As iron deficiency progresses, development of anemia precedes appearance of morphologic changes in blood, although some cells may be smaller and paler than normal; serum iron concentration is usually low at this time, but it may be normal. With advanced iron depletion, classic changes of hypochromic, microcytic, hypoferremic anemia become manifest. (Reproduced with permission from Lichtman’s Atlas of Hematology, www.accessmedicine.com.)
As the body becomes depleted of iron, changes occur in many tissues. Hemosiderin and ferritin virtually disappear from marrow and other storage sites. Hemoglobin synthesis in the marrow decreases, first as a result of fewer erythroblasts,70 but eventually also per erythroblast if iron deficiency becomes more severe, resulting in hemoglobin-deficient erythrocytes. The concentration of many other iron-containing proteins is affected, often in an organ-specific manner.71 Studies in laboratory animals on defined iron-deficient diets are most informative about this process, because human iron deficiency is often confounded by other forms of malnutrition. In such models of severe (pure) iron deficiency, skeletal muscle myoglobin is mildly depleted but cardiac myoglobin is not. Cytochromes and other mitochondrial ferroproteins are depleted but selectively so. Since these classical studies were performed, it has become apparent that the synthesis of many ferroproteins is regulated in an iron-dependent manner, mainly via the iron-responsive element (IRE)/iron-regulatory protein (IRP) system (Chap. 42). The changes in iron-containing proteins may thus be adaptive,72 to allow the survival of the organism until more iron becomes available.
Muscular Function and Exercise Tolerance
Decrements in high-intensity exercise performance can be detected even during nonanemic iron deficiency,63 and worsen with increasing anemia.73 The limitation of high-level exercise by oxygen delivery, and, therefore, hemoglobin content of blood, is well known, and has given rise to surreptitious blood doping and erythropoietin abuse by some athletes. The impairment of performance during nonanemic iron deficiency consists of decreased spontaneous activity (seen in humans and in animal models) and decreased ventilatory threshold, that is, the point at which ventilation starts increasing more rapidly than oxygen consumption.74 Other deficits that have been reported include decreased endurance and increased muscle fatigue. The biochemical basis of the deficits associated with nonanemic iron deficiency is not well understood but is attributed to the depletion of iron-containing mitochondrial proteins that are involved in energy metabolism.63 The condition is reversible with iron supplementation.
Iron deficiency is associated with both developmental abnormalities in children and with restless leg syndrome in adults, but in neither case has iron deficiency been established as the primary cause.75,76 The substantia nigra is a particularly iron-rich region of the brain and contains dopaminergic neurons that are suspected of involvement in restless leg syndrome. In mouse models of iron deficiency, iron depletion of the substantia nigra is highly strain-dependent,77 suggesting that iron deficiency and as yet incompletely characterized genetic variations may cooperate in the pathogenesis of restless leg syndrome by allowing the depletion of iron from susceptible brain regions involved in dopaminergic signaling.78
Host Defense and Inflammation
In multiple publications, iron deficiency has been reported to impair various immune functions, but the effects appear minor and inconsistent.79 Perhaps surprisingly, the evidence for a narrowly protective and proinflammatory effect of iron deficiency appears stronger. Iron deficiency decreases the risk and severity of malaria,80,81 and iron supplementation may have the opposite effect, especially when not targeted to patients with iron deficiency.52,82 The mechanism of this effect is of great interest but not yet well understood.83 There are some indications that iron deficiency may have a proinflammatory effect. In a mouse model, iron deficiency potentiated the systemic effect of lipopolysaccharide in a hepcidin-dependent manner,84 and in a mouse model of asthma, iron deficiency promoted allergic inflammation.85
Iron-deficient children have been reported to suffer from growth retardation but it is difficult to isolate the effect of iron deficiency from other nutritional and environmental causes of stunting. Two comprehensive analyses of randomized controlled trials did not detect an effect on growth of iron supplementation alone.86,87 Decreased thermoregulation in response to cold exposure is seen in both humans and laboratory models.88 It has been attributed to the conflicting effects on blood flow of decreased oxygen content of blood and need to minimize heat loss, as well as the effect of iron deficiency on thyroid function.
Severe iron deficiency may lead to histologic changes in various organs. The rapidly proliferating cells of the upper part of the alimentary tract seem particularly susceptible to the effect of iron deficiency. There may be atrophy of the mucosa of the tongue and esophagus,89 stomach,90 and small intestine.91 The epithelium of the lateral margins of the tongue is reduced in thickness despite increase in the progenitor compartment. This thinning presumably reflects accelerated exfoliation of epithelial cells.92 Buccal mucosa has shown thinning and keratinization of epithelium and increased mitotic activity.93,94 However, light microscopic and electron microscopic examination of exfoliated oral mucosal cells showed no aberrations in morphology of nuclei or cytoplasm of the cells of patients with iron-deficiency anemia.95 In iron-deficiency anemia resulting from idiopathic pulmonary hemosiderosis, characteristic pathologic changes are found in the lungs, including intense deposition of iron in the littoral cells of the alveoli and interstitial fibrosis.45
Widening of diploic spaces of bones, particularly those of the skull and hands,96,97,98 may be a consequence of chronic iron deficiency beginning in infancy. In the skull, this is of the same character as in thalassemia, except that in β-thalassemia major there is maxillary hypertrophy, whereas in severe iron-deficiency anemia maxillary growth and pneumatization are normal.
Clinical Manifestations of Anemia
The anemia in iron-deficient patients can be very severe, with blood hemoglobin levels as low as less than 4 g/dL being encountered in some patients. Severe iron-deficiency anemia is associated with all of the various symptoms of anemia, resulting from hypoxia and the body’s response to hypoxia, as described in Chap. 32. Thus, tachycardia with palpitations and pounding in the ears, headache, light-headedness, and even angina pectoris, may all occur in patients who are severely anemic.
Clinical Manifestations That May Be Unrelated to Anemia
The clinical features of iron deficiency encompass nonhematologic effects and symptoms caused by the anemia itself. A number of controlled studies show that various manifestations of iron deficiency can occur in individuals whose hemoglobin is within the accepted normal range.
Decreased Work Performance
Objective measurements of work performance and studies using O2 consumption as an index of work performance have given contradictory results, but a comprehensive review led to the conclusion that severe iron-deficiency anemia (hemoglobin <8 g/dL) and mild iron deficiency anemia (hemoglobin between 8 and 12 g/dL) led to decreased work performance, primarily as estimated by peak oxygen consumption (VO2max) measurements, but the evidence that nonanemic iron deficiency had such an effect was less convincing.99 However, in athletes with low ferritin levels but normal hemoglobin levels, iron-supplemented subjects showed an increased VO2max without a change in their red cell mass, and in other studies nonanemic subjects treated with iron showed improved performance and/or VO2max.63
Infant and Childhood Development
It has been proposed that iron deficiency in infants and children is associated with poor attention span, poor response to sensory stimuli, and retarded behavioral and developmental achievement even in the absence of anemia. The causality of these associations is confounded by other coexisting nutritional deficits and socioeconomic deprivation, so reversibility by iron supplementation would be important in establishing causality. However, in systematic meta-analyses, iron supplementation had weak or no effects on these deficits.86,100,101,102
It has been speculated that there is a relationship between restless legs syndrome, Tourette syndrome, and attention deficit hyperactivity disorder and that iron deficiency contributes to their pathophysiology. Restless legs syndrome, a common nocturnal problem, especially in the elderly, is associated with iron deficiency and is reported to improve on iron therapy, but the beneficial effects are inconsistent and not well predicted by blood ferritin or transferrin saturation.75,103,104 In children there may be a relationship between iron deficiency and attention deficit hyperactivity disorder, but the association is inconsistent.105
Other Neurologic Symptoms
Breath holding in children, headaches, and paresthesias have been attributed to iron deficiency but there are no controlled studies to support these impressions. Anecdotal reports of intracranial hypertension with papilledema are buttressed by apparent response to iron therapy.106,107,108,109 Stroke in children and in adults, possibly triggered by thrombocytosis, is associated with iron-deficiency anemia.110,111,112,113,114
Oral and Nasopharyngeal Symptoms
Burning of the tongue has also been described anecdotally in many accounts of iron deficiency, and although this symptom has been observed to diminish with treatment, no controlled studies have been performed. The tongue symptoms could be a result of concurrent pyridoxine deficiency. Although iron deficiency has been proposed as a cause of atrophic rhinitis, the evidence for this is weak.
In the laryngopharynx, mucosal atrophy may lead to web formation in the postcricoid region, thereby giving rise to dysphagia (Paterson-Kelly also known as Plummer-Vinson syndrome).115 If these alterations are of long duration, they may lead to pharyngeal carcinoma. Although it has been generally thought that these changes are secondary to longstanding iron deficiency, this mechanism is not universally accepted. The frequency of the condition is considered to have decreased considerably, and it is remarkably rare in many parts of the world where iron deficiency is common.
The craving to eat unusual substances, for example, dirt, clay, ice, laundry starch, salt, cardboard, and hair, is a well-documented manifestation of iron deficiency and is usually cured promptly by iron therapy.116,117,118,119
Although the association of hair loss with iron deficiency is controversial,120 low ferritin levels were a risk factor for hair loss in a large multivariate analysis.121 Remarkably, hair loss sparing the face (“mask mouse”) is a sign of iron deficiency in mice.122
The physical findings in iron-deficiency anemia include pallor, glossitis (smooth, red tongue), stomatitis, and angular cheilitis. Koilonychia (spoon nails), once a common finding, is now encountered rarely. Retinal hemorrhages and exudates may be seen in severely anemic patients (e.g., hemoglobin concentration of <5 g/dL). Splenomegaly has occasionally been attributed to iron-deficiency anemia, but when it occurs, it is probably from other causes.
In severe, uncomplicated iron-deficiency anemia, the erythrocytes are hypochromic and microcytic; the plasma iron concentration is diminished; the iron-binding capacity is increased; the serum ferritin concentration is low; the serum transferrin receptor (TfR) and erythrocyte zinc protoporphyrin concentrations are increased; and the marrow is depleted of stainable iron. However, the classic combination of laboratory findings occurs consistently only when iron-deficiency anemia is far advanced, when there are no complicating factors such as infection or malignant neoplasms, and when there has not been previous therapy with transfusions or parenteral iron.
Anisocytosis is the earliest recognizable morphologic change of erythrocytes in iron-deficiency anemia (Fig. 43–2).123 The anisocytosis is typically accompanied by mild ovalocytosis. As the iron deficiency worsens, a mild normochromic, normocytic anemia often develops. With further progression, hemoglobin concentration, erythrocyte count, mean corpuscular volume (MCV), and mean erythrocyte hemoglobin content all decline together.124,125 As the indices change the erythrocytes appear microcytic and hypochromic on stained blood films. Target cells may sometimes be present. Elongated hypochromic elliptocytes may be seen, in which the long sides are nearly parallel. Such cells have been called “pencil cells,” although they more nearly resemble cigars in shape. The red cell indices are consistently abnormal in adults only when iron-deficiency anemia is moderate or severe (e.g., in males with hemoglobin concentrations <12 g/dL or in women with hemoglobin concentrations <10 g/dL) (Fig. 43–3). The distribution of erythrocyte volume (e.g., red cell distribution width [RDW]) is usually increased in established iron-deficiency anemia. The RDW is reported often as the coefficient of variation (in percent) of erythrocyte volume (see “Differential Diagnosis” below).
Variability in morphologic diagnosis of iron-deficiency anemia from blood film. As in all deficiency states leading to anemia, the blood film morphology and blood cell changes are a function of the severity of the deficiency. A. Normal blood film. Normocytic-normochromic red cells with normal shape. B. Mild iron deficiency. Serum iron, ferritin, and transferring saturation were consistent with mild iron deficiency. Cannot discern if mean red cell size has decreased. There may be a few red cells that have larger central pallor, but that is arguable. A few cells have oval or elliptical shape. C. Severe iron deficiency. Serum iron, ferritin, and transferring saturation were consistent with severe iron deficiency. Note obvious increase in overtly hypochromic cells and higher frequency of microcytes. (Reproduced with permission from Lichtman’s Atlas of Hematology, www.accessmedicine.com.)
Erythrocyte indices in iron-deficiency anemia of adults; data obtained with Coulter Counter, Model S. Normal ranges of indices observed in approximately 500 healthy adults using the same instrument are indicated by shading. The dashed line in the top panel indicates the more widely accepted lower normal limit of mean corpuscular hemoglobin concentrations (MCHCs) stated in this text. (Top) Correlation between venous blood hemoglobin concentration and MCHC. More than half of 62 patients with iron-deficiency anemia had MCHC values clearly in the normal range. (Bottom) Correlation between venous blood hemoglobin concentrations and mean corpuscular volume (MCV). Nearly 70 percent of cases exhibited distinct microcytosis. Thus when indices are determined by automated cell-counting methods, the MCV is much more sensitive than is the MCHC in detecting changes of iron deficiency. However, at least 30 percent of cases of iron-deficiency anemia will be misdiagnosed if physicians rely on the erythrocyte indices. (Data from Klee GG: Decision Rules for Accelerated Hematology Laboratory Investigation: Thesis, University of Minnesota.)
Leukopenia has been found in some patients with iron-deficiency anemia, but the overall distribution of leukocyte counts in iron-deficient patients seems to be approximately normal.
Both thrombocytopenia126 and, more commonly, thrombocytosis127 have been associated with iron deficiency. Platelet abnormalities correct with iron therapy. Thrombotic complications of iron deficiency have been reported but are very rare.128 The etiology of either abnormality is not known. Low-iron-diet-induced iron-deficiency anemia developed in a rat model within 2 weeks, and this was accompanied by sustained 50 percent increase in platelet count with increased platelet size but without significant changes in known megakaryocyte growth factors (thrombopoietin, IL-6 or IL-11). It has been suggested that high erythropoietin levels may stimulate thrombopoietin receptors because the two hematopoietic factors are structurally related, but this does not seem to be the case.129
Reticulocyte count is often mildly increased,130 a finding consistent with the increased erythroid activity of the marrow (see “Marrow” below).
Because most of the iron in the body is normally in erythrocytes, and iron is not excreted, decrease in erythrocyte mass generally results in increased storage iron. Iron-deficiency anemia is the exception, as iron stores are depleted before the red cell mass is compromised. Thus, evaluation of iron stores should be a sensitive and usually reliable means for the differentiation between iron-deficiency anemia and all other anemias. Decreased or absent hemosiderin in the marrow is characteristic of iron deficiency, and is readily evaluated after staining by the simple Prussian blue method. Stored iron in the macrophages of the marrow can be seen in marrow spicules in marrow sections, or in marrow aspirate films. Iron granules, normally found in the cytoplasm of approximately 30 percent of erythroblasts, become rare but may not be entirely absent.
Evaluation of the amount of iron in marrow macrophages has long been considered the “gold standard” for the diagnosis of iron deficiency. There are, however, technical barriers to the accurate histochemical determination of marrow iron. First, an invasive procedure, marrow aspiration, is required. Second, the differentiation of iron within macrophages from artifacts takes experience and skill. In one study only 74 of 108 cases had been accurately reported.131 Moreover, misleading results may be obtained in patients who have been transfused or who have been treated with parenteral iron.132 The marrow of such patients may contain normal, or even increased, quantities of stainable iron in the face of typical iron-responsive iron-deficiency anemia. In such patients, iron that is seen on marrow examination is not readily available for erythropoiesis. As serum markers of iron deficiency became widely available, the reasons for the primacy of marrow iron estimation have been questioned.133
The serum iron concentration is usually low in untreated iron-deficiency anemia, but may rarely be normal.125,134,135 Iron in blood plasma turns over every few hours and constitutes less than 0.1 percent of total body iron in adults, so iron concentrations are readily perturbed by transient changes in iron supply or demand. Physiologically, the serum iron concentration has a diurnal rhythm; it decreases in late afternoon and evening, reaching a nadir near 9 pm and increases to its maximum between 7 and 10 am. This effect is rarely of sufficient magnitude to influence diagnosis.136 Serum iron levels decrease at about the time of menstrual bleeding137,138 regardless of whether the bleeding is physiologic or induced by withdrawal of contraceptive hormonal preparations.
Importantly, the serum iron concentration is reduced in the presence of either acute or chronic inflammatory processes139 or malignancy140 and following acute myocardial infarction.141,142 The serum iron concentration under these circumstances may be decreased sufficiently to suggest iron deficiency. Conversely, during chemotherapy of malignancy, the serum iron concentration may be quite elevated, as cytotoxic effects of the drugs on erythroblasts inhibit erythropoiesis and related iron uptake by erythroblasts. This effect is observed from the third to the seventh day after inception of chemotherapy of a variety of tumors.143
Normal or high concentrations of serum iron are commonly observed even in patients with iron-deficiency anemia if such patients receive iron medication before blood is drawn for these measurements. Even multivitamin preparations, which commonly contain approximately 18 mg of elemental iron per tablet, can result in this effect. Oral iron medication should be withheld for 24 hours before blood samples are obtained. Parenteral injection of iron dextran may result in a very high serum iron concentration (e.g., 500 to 1000 mcg/dL), at least with some methods,144 for several weeks. The elevation of serum iron levels after infusion of sodium ferric gluconate or iron sucrose is of much shorter duration.145
Iron-Binding Capacity and Transferrin Saturation
The iron-binding capacity is a measure of the amount of transferrin in circulating blood. Normally, there is enough transferrin present in 100 mL serum to bind 4.4 to 8.0 μmol (250 to 450 mcg) of iron; because the normal serum iron concentration is approximately 1.8 μmol/dL (100 mcg/dL), transferrin may be found to be approximately one-third saturated with iron. The unsaturated or latent iron-binding capacity (UIBC) is easily measured with radioactive iron or by spectrophotometric techniques. The sum of the UIBC and the plasma iron represents total iron-binding capacity (TIBC). TIBC can also be measured directly. In iron-deficiency anemia, UIBC and TIBC are often increased and serum iron concentrations are decreased so that transferrin saturation of 15 percent or less is usually found. Because transferrin concentration and TIBC are decreased during inflammation, a normal value for transferrin saturation often accompanies a low serum iron concentration in the anemia of chronic inflammation.
Serum ferritin, secreted mainly by macrophages146 and hepatocytes, contains relatively little iron, yet serum ferritin concentration empirically correlates with total-body iron stores,147 for reasons that are still obscure. Serum ferritin concentrations of 10 mcg/L or less are characteristic of iron-deficiency anemia. In iron deficiency without anemia, serum ferritin concentration is typically in the range of 10 to 20 mcg/L. An increase in serum ferritin concentration occurs in inflammatory disorders, such as rheumatoid arthritis, in chronic renal disease, and in malignancies.148 When one of these conditions coexists with iron deficiency, as they often do, the serum ferritin concentration is commonly in the normal range; interpretation of results of this assay then becomes difficult. In patients with rheumatoid arthritis who are anemic, some suggest that concomitant iron deficiency may be suspected when the serum ferritin concentration is less than 60 mcg/L,149 but such empiric guidelines are unlikely to apply to the full spectrum of severity of inflammation. Increased serum ferritin concentrations are also characteristic of some malignancies, as well as of acute and chronic liver disease and chronic renal failure.150,151,152,153 In Gaucher disease, juvenile rheumatoid arthritis, and various macrophage activation syndromes, and in ferroportin disease characterized by massive iron loading of macrophages, the serum ferritin concentration is commonly in the range of thousands of mcg/L and may mask iron deficiency.154,155,156,157,158
Erythrocyte Zinc Protoporphyrin
Erythrocyte protoporphyrin, principally zinc protoporphyrin, is increased in disorders of heme synthesis, including iron deficiency, lead poisoning, and sideroblastic anemias, as well as other conditions.159,160,161 This assay analyzes the fluorescence of erythrocytes and uses small blood samples. It is quite sensitive in the diagnosis of iron deficiency and practical for large-scale screening programs designed to identify children with either iron deficiency or lead poisoning.58,159 It does not differentiate between iron deficiency and anemia that accompanies inflammatory or malignant processes.162
Serum Transferrin Receptor
The role of TfR in transporting transferrin iron into cells is described in Chap. 42 section “Transport of Iron”. The circulating receptor is a truncated form of the cellular receptor, lacking the transmembrane and cytoplasmic domains of the cellular receptor. It circulates bound to transferrin. Sensitive immunologic methods can detect approximately 5 mg/L of receptor in serum. The levels of circulating TfR mirror the amount of cellular receptor, and therefore are proportional to the number of erythroblasts expressing the receptor. Because receptor synthesis is greatly increased when cells lack iron, the amount of the circulating receptor increases in iron deficiency.163,164 In anemia of inflammation, the synthesis of the TfR is suppressed by cytokines and this negates the opposing stimulatory effect of iron restriction, resulting in a lower serum TfR concentration than in pure iron deficiency.165 This test for iron deficiency has gradually come into clinical use, but the methodology has not yet been standardized, making laboratory-to-laboratory comparisons difficult. A method for performing reproducible assays for the soluble TfR has been standardized.166 Like the serum ferritin and serum iron, serum TfR assay results may be confounded by poorly understood variations in patients with malignancies; in patients in whom the serum TfR concentration is reduced; and in patients with asymptomatic malaria or thalassemia trait,167,168 in whom, in the absence of iron deficiency, it is increased. The ratio of serum TfR to serum ferritin seems to be a useful but not infallible reflection of body iron stores.169 Moreover, several studies show that the soluble transferrin index calculated as a ratio of the serum TfR/log ferritin (TfR-F Index) may be superior to other means for detection of iron deficiency.170,171,172
Reticulocyte Hemoglobin Content and Other Novel Erythrocyte Indices
Some automated hematology instruments offer a method for diagnosis of iron deficiency using an assay of hemoglobin content within reticulocytes. This parameter is an indicator of iron restriction of hemoglobin synthesis during 3 to 4 days prior to the test.173,174 Percent hypochromic erythrocytes offers a longer term assessment of iron restriction during the preceding few months.173,175
Iron-deficiency anemia is characterized by many abnormal laboratory features. Because none of these are unique, a small deviation from normal will detect most cases of iron deficiency (high sensitivity), but also falsely identify non–iron-deficient subjects as being iron deficient (low specificity). On the other hand, a large deviation from normal will exclude most nondeficient patients (high specificity), but miss many iron-deficient subjects (low sensitivity). This tradeoff is shown graphically in so-called receiver operator characteristic curves. These curves are constructed by plotting the sensitivity against the false-positive rate (1 − specificity) at various values of the analyte. Figure 43–4 shows receiver operator characteristic curve for some tests for iron deficiency. The situation is complicated in the case of iron deficiency by the fact that the diagnostic problem faced by the physician is not one of differentiating a patient with iron-deficiency anemia from a normal person, but rather from a patient who has an anemia with a different etiology. It is partly for this reason that a simple algorithm for the diagnosis of iron deficiency does not exist. In a severely anemic patient, microcytosis would have very high specificity and high sensitivity compared to normal, but compared to a patient with thalassemia the specificity would be very low. Similarly, a low serum ferritin level is an excellent test in the general population, but it has relatively little value in patients with chronic renal disease. Another problem that is inherent in evaluating diagnostic tests for iron deficiency is the standard that is applied to decide who is iron deficient and who is not. Marrow iron has served as one “gold standard” but has limitations, as discussed earlier (see “Marrow” earlier). Alternatively, the response to iron therapy serves as a powerful indicator of whose anemia is actually a result of a deficiency of iron. Here, too, there are limitations, in that some iron-deficient patients may fail to respond adequately because of factors such as infection. Lacking an absolute test for iron deficiency, the ability of the physician to use judgment relevant to the particular patient’s circumstances is of paramount importance.
Two receiver operator curves. As the specificity increases the sensitivity decreases. The receiver-operator properties of serum ferritin are far from ideal. When the specificity is high (to the left on the abscissa) the sensitivity is low; only when the specificity is low is the sensitivity adequate. The curve that would be obtained with a nearly ideal test for iron deficiency gives high specificity and high sensitivity. In the curve shown, a cutoff value could be found that allows one to identify 75 percent of patients with iron deficiency with a specificity of greater than 90 percent. Unfortunately, no such test exists.
The forms of anemia that must be distinguished from iron-deficiency anemia most frequently include those of thalassemia minor, chronic inflammatory disease, malignancy, chronic liver disease, and chronic renal disease. It is the microcytic anemias that are most likely to be confused with iron deficiency. These include other conditions in which hemoglobin synthesis is impaired,176 including thalassemias and thalassemia traits, drug- or toxin-induced impairments of heme synthesis, sideroblastic anemias (Chap. 59), and very rare defects in the delivery of iron to erythrocytes or erythrocyte iron uptake and utilization (Table 43–2).
Table 43–2.Microcytic Disorders Other Than Iron Deficiency ||Download (.pdf) Table 43–2. Microcytic Disorders Other Than Iron Deficiency
|Mechanisms ||Diseases |
|Impaired globin chain synthesis or highly unstable hemoglobin ||β-Thalassemia or trait, α-thalassemia minima or minor, hemoglobin H, hemoglobin E or trait, combinations of above |
|Drugs or toxins that inhibit heme synthesis ||Lead, isoniazid, pyrazinamide, sirolimus |
|Disorders that impair heme synthesis directly or by decreased iron delivery to erythroblasts, or decreased uptake or utilization of iron by erythroblasts ||sideroblastic anemias, erythropoietic porphyrias, atransferrinemia,203 aceruloplasminemia,284 DMT-1 mutations,207 STEAP3 deficiency285 |
In many parts of the world, and in many communities of North America, the frequency of β-thalassemia minor is second only to that of iron deficiency as a cause of hypochromic microcytic anemia (Chap. 48). In African Americans, homozygosity for α-thalassemia-2, that is the state in which only single α-globin gene is present on each chromosome, is a common cause of microcytosis. Approximately 3 percent of African Americans are homozygous for α-thalassemia-2. The condition is associated with only a very modest lowering of the blood hemoglobin level.177 Heterozygotes may also have microcytosis, although usually they are hematologically normal. Among persons of Mediterranean ancestry both α- and β-thalassemia are very prevalent, particularly the latter. Among Asians, particularly in those from Southeast Asia, β-thalassemia minor, α-thalassemia minor, and hemoglobin E trait, all occur frequently. All are characterized by microcytosis, and none can be distinguished reliably from the others on the basis of erythrocyte morphology or erythrocyte indices alone. In each of these conditions there may be only mild to moderate microcytosis without any other distinctive changes. However, in the majority of patients with α- or β-thalassemia minor, hemoglobin Lepore trait, and hemoglobin E trait, the erythrocyte count is greater than 5 × 1012/L (5,000,000/μL), despite low hemoglobin concentration.178,179 Homozygous hemoglobin E is also characterized by marked hypochromia, microcytosis, abundant target cells, and elevated erythrocyte count, but usually not by more than minimal anemia (Chap. 49).178
In contrast to the findings in these hemoglobinopathies, erythrocyte counts of 5 × 1012/L (5,000,000/μL) or higher are relatively uncommon among adults with iron-deficiency anemia.180 However, erythrocytosis may be seen in children with iron-deficiency anemia or in polycythemia vera patients who have become iron deficient following hemorrhage or therapeutic phlebotomy. Consequently, while the mean MCV is almost always reduced in α- or β-thalassemia minor and in homozygous hemoglobin E, with values of 60 to 70 fL being the rule, values this low are seen only in severe iron-deficiency anemia. In hemoglobin Lepore trait and hemoglobin E trait, only minimal microcytosis is observed.178,179,181 Algorithmic rules based on red cell counts, MCV or RDW are not sufficiently reliable for distinguishing iron deficiency from thalassemia in populations with high prevalence of iron deficiency compared to thalassemias.
Mild reticulocytosis, polychromatophilia, and basophilic stippling are more likely to be encountered in β-thalassemia minor, δβ-thalassemia minor, and hemoglobin Lepore trait than in iron-deficiency anemia, but may be absent in these disorders. The serum iron concentration is usually normal or increased in thalassemic syndromes and is usually low in iron-deficiency anemia. Similarly, examination of marrow iron stores helps to differentiate these disorders. The presence of β-thalassemia trait is substantiated by the demonstration of increased proportions of hemoglobin A2 and F, or by the presence on electrophoresis of hemoglobin H or Lepore (Chap. 48). At present, the diagnosis of α-thalassemia minor is usually made on the basis of exclusion of other causes of microcytosis, but it can be confirmed by direct demonstration of mutations in α-globin genes by DNA-based techniques.
Iron deficiency may mask concurrent thalassemia. The amounts of both hemoglobin A2 and hemoglobin H are diminished disproportionately to the reduction in hemoglobin A in the presence of iron deficiency182 (Chap. 46); however, usually the hemoglobin A2 level remains above the normal range.
Anemia of Inflammation (Anemia of Chronic Disease)
The anemia of inflammation (Chap. 37) is usually normochromic and normocytic, but hypochromic microcytic anemia occurs in 20 to 30 percent of patients with chronic infections or malignancies.139 Thus these disorders cannot be distinguished from iron-deficiency anemia by examination of the blood film. Furthermore, the serum iron concentration is usually decreased in these disorders,139 sometimes severely. In uncomplicated iron deficiency, the TIBC is usually increased, whereas in inflammatory and neoplastic diseases it is commonly decreased, but there is considerable overlap among TIBC values of normal subjects, those with iron-deficiency anemia, and those with chronic inflammatory diseases.
Transferrin saturation may be normal in iron-deficiency anemia, and, conversely, low saturation is sometimes observed in chronic inflammation. However, circulating soluble TfRs increase in iron deficiency but not in the anemia of inflammation.170 The serum ferritin level is usually diminished in iron deficiency, but it is generally increased in chronic inflammatory and neoplastic disorders.147 Measurement of the ratio of soluble TfR to ferritin has been found to be useful in distinguishing the anemia of chronic inflammation from that of iron deficiency,170 but meta-analysis of the relevant clinical studies suggested that the ratio may not be better than soluble TfR assay alone at discriminating between iron deficiency anemia and anemia of inflammation.183 Examination of the marrow for stainable iron is invasive and requires skilled reading but may be helpful in an occasional patient. Iron staining of marrow macrophages is greatly decreased in amount or absent in iron deficiency anemia and normal or increased in the other disorders. Low serum hepcidin concentrations are characteristic of iron deficiency and high serum hepcidin indicates anemia of inflammation, but this assay is not yet clinically available and its clinical value is unknown. As would be expected from the inhibitory effect of hepcidin on the absorption of iron, high hepcidin levels predict a poor response to oral iron therapy184 and low incorporation of dietary iron into erythrocytes.185
Anemia of Chronic Liver Disease
The erythrocytes in the blood film from patients with chronic liver disease may be normochromic and normocytic, macrocytic, or hypochromic. Target cells are frequently present in large numbers. Because the blood film in iron-deficiency anemia may also display these features, differential diagnosis must be based on other observations. Low serum ferritin levels are useful in detecting iron deficiency in the setting of cirrhosis,186 but normal or even increased serum ferritin does not exclude iron deficiency, especially in the presence of active liver injury.186,187
Anemia of Chronic Renal Disease
Iron deficiency is frequent in patients with chronic renal disease (Chap. 37). Iron-deficiency anemia is particularly difficult to diagnose in patients with chronic renal disease (Chap. 37) where iron delivery to the marrow is inadequate because of coexisting inflammation and because of increased demands from pulsatile erythropoiesis as a result of erythropoiesis-stimulating agents. Because the problem is fairly common, and perhaps because of interest in identifying those patients who can benefit from iron therapy and decreasing their use of erythropoiesis-stimulating agents, a large number of studies have been done to determine the best way to diagnose iron deficiency in patients undergoing extracorporeal dialysis. The diagnostic problem is further complicated by the common occurrence of “functional iron deficiency,”188 that is, a state in which iron stores are adequate but iron delivery to the marrow is insufficient to meet the increased kinetic requirements of erythropoiesis stimulated by intermittent use of erythropoietin and related agents. If response to intravenous iron therapy is used to diagnose iron deficiency, even many patients with abnormally high ferritins will be iron deficient.150,189 High serum ferritins do not preclude a response to IV iron, not even in patients with chronic kidney disease who are not on hemodialysis.190 Although measurements of reticulocyte hemoglobin and percent of hypochromic erythrocytes show promise as markers of response to iron therapy, there is insufficient evidence that any individual biomarker or combination of biomarkers can reliably predict the response to iron treatment in chronic kidney disease.191
Anemia of Hemolytic Disease
Hemolytic disease can usually be distinguished from iron-deficiency anemia on the basis of the blood film. The marked polychromatophilia, spherocytosis, schistocytes, Heinz bodies, basophilic stippling, and other morphologic features characteristic of various types of hemolysis usually are not seen in iron-deficiency anemia. Furthermore, reticulocytosis is usually marked in hemolytic disorders but minimal or absent in iron-deficiency anemia. However, there are some outstanding exceptions to these generally valid principles.
In unstable hemoglobin disorders, such as hemoglobin H disease or hemoglobin Köln disease, erythrocytic hypochromia may be pronounced. In these disorders, there is moderate reticulocytosis, which helps to differentiate them from iron-deficiency anemia. The serum iron concentration is normal or increased. Chapter 49 discusses the detection of unstable hemoglobins.
When there is chronic intravascular hemolysis, erythrocytes in the blood film may display marked morphologic abnormalities, such as burr cells and schizocytes. Yet because of loss of iron in the urine, iron deficiency may be the dominant cause of the resulting anemia. Evaluation of iron content in marrow aspirates or measurement of serum iron concentration and TIBC may clarify the diagnosis in this form of anemia.
Hypoplastic and Aplastic Anemia
In their early phases, these disorders cannot reliably be differentiated from mild iron-deficiency anemia on the basis of erythrocyte morphology alone (Chap. 35). The reticulocyte count is generally less than 0.5 percent in hypoplastic or aplastic anemia. The presence of neutropenia and thrombocytopenia suggests a diagnosis of aplastic anemia, but mild neutropenia may also occur in iron-deficiency anemia. The serum iron concentration is usually increased in aplastic anemia; and the percentage transferrin saturation is then elevated. Marrow aspiration may produce scant material for cytologic study, and marrow biopsy may be necessary. An iron stain usually reveals increased amounts of hemosiderin in aplastic or hypoplastic anemia. However, if chronic bleeding has occurred, for example, as a consequence of thrombocytopenia, iron stores may be depleted.
In this heterogeneous group of disorders (Chap. 59), the blood findings often simulate those of iron-deficiency anemia. Reticulocytosis is usually absent, and the serum iron concentration and serum ferritin is generally normal or increased. Marrow examination shows characteristic ring sideroblasts and increased amounts of stainable iron.
Congenital Dyserythropoietic Anemia
In the rare congenital dyserythropoietic anemias (Chap. 39), erythrocyte morphologic abnormalities may resemble those of iron deficiency or thalassemia (Chap. 48). In general, in congenital dyserythropoietic anemias, poikilocytosis is very striking and occurs with less reduction in MCV than in iron deficiency or thalassemias. Often, however, such cases are believed to be thalassemic until the marrow is examined.
In pernicious anemia and other types of megaloblastic anemia (Chap. 41), the blood film usually shows changes sufficiently distinctive that there is little difficulty in differential diagnosis. One potential source of error is the change in serum iron concentration that occurs after therapy. In the patient with pernicious anemia or folic acid deficiency, early after starting treatment, the serum iron concentration decreases markedly as iron is used rapidly for hemoglobin synthesis.192 Thus the finding of a low serum iron concentration in such circumstances should not be taken as evidence of iron deficiency. Iron-deficiency anemia and anemia as a consequence of folic acid or vitamin B12 deficiency may coexist. During the course of treatment, with the rapid increase in the number of red cells, the typical manifestations of severe iron deficiency may develop. The mixture of microcytic-hypochromic and normocytic-normochromic cells has been called dimorphic anemia (see “Coexisting Microcytic Anemia” in Chap. 41).
The anemia of severe hypothyroidism (myxedema; Chap. 38) is usually normochromic and normocytic and may be accompanied by mild-to-moderate depression of serum iron concentration. Marrow examination may be required to determine whether iron deficiency is present, especially as iron deficiency often complicates myxedema because of menorrhagia, which is common in this disorder.
In the final analysis, the response to iron therapy is the proof of correctness of diagnosis of iron-deficiency anemia. Furthermore, some physicians or patients may not have access to all the techniques described for diagnosis of iron-deficiency anemia. In this event, the patient’s response to therapy may become a primary diagnostic measure. Iron administration in such a therapeutic trial is usually by the oral route, but intravenous iron can be used if there is evidence or strong suspicion of coexisting inflammation, iron malabsorption, or intolerance of oral iron preparations. A therapeutic trial under any circumstances should be followed carefully. If the cause of anemia is iron deficiency, adequate iron therapy should result in reticulocytosis with a peak occurring after 1 to 2 weeks of therapy, although if anemia is mild, the reticulocyte response may be minimal. A significant increase in the hemoglobin concentration of the blood should be evident 3 to 4 weeks later, and the hemoglobin concentration should attain a normal value within 2 to 4 months. Unless there is evidence of continued substantial blood loss, the absence of response to oral or, when appropriate, parenteral iron must be taken as evidence that iron deficiency is not the cause of anemia. Iron therapy should be discontinued and another cause for the anemia sought.
Special Studies to Delineate the Cause of Iron Deficiency
The physician who establishes a diagnosis of iron deficiency resulting from blood loss has the obligation to determine the site and cause of hemorrhage. Examination for fecal occult blood is particularly helpful in determining what additional studies should be carried out. Specimens should be examined on at least 3 days, because bleeding may be intermittent. Occasionally, it is helpful to label the patient’s erythrocytes with chromium-51 (51Cr) sodium chromate and to determine quantitatively the amount of blood lost daily. When there is reason to believe that bleeding is from the gastrointestinal tract, roentgenographic and other imaging studies and endoscopic investigation are indicated. The other imaging studies often include gastroscopy, esophagoscopy, colonoscopy, and capsule endoscopy, and, rarely, angiography or scintigraphic studies. Numerous clinical studies indicate that intensive investigation of patients, particularly men and postmenopausal women, reveals unexpected bleeding lesions, many of which are curable or treatable.10,193 H. pylori infection should be sought, particularly in patients who are iron deficient but who do not seem to respond to therapy.14,15 An iron stain of sputum may reveal hemosiderin-laden macrophages when there is intrapulmonary bleeding.
Once it has been established that a patient is deficient in iron, replacement therapy should be instituted. Iron may be administered orally, as simple iron salts; parenterally, as an iron-carbohydrate complex; or, very rarely, as a blood transfusion. In general, the oral route is preferred, but the intravenous route is increasingly used because of the improved safety and convenience of new parenteral iron preparations. In most patients, iron-deficiency anemia is a disorder of long duration and slow progression, and restoration of normal hemoglobin is not urgent unless the patient suffers from acute cardiac problems, in which case blood transfusion is appropriate. There is usually time to wait for normal mechanisms of erythropoiesis to respond to the body’s needs and for gradual adjustment of the cardiovascular system to reexpansion of the total circulating erythrocyte volume.
The patient should be encouraged to eat a diversified diet supplying all nutritional requirements. Nonetheless, it must be emphasized that neither meat nor any other dietary article contains enough iron to be useful therapeutically. Meat contains small amounts of myoglobin and hemoglobin and insignificant amounts of iron in other proteins. Although heme iron is better absorbed than inorganic iron, the quantity of heme iron in meat is actually quite small. In fact, an average (3-ounce) serving of steak provides only about 3 mg of iron, that is, the equivalent of only 3 mL of packed erythrocytes. Provision of sufficient dietary iron to permit a maximal rate or recovery from iron-deficiency anemia might require a daily intake of at least 10 pounds of steak. For these and other reasons, medicinal iron is much superior to dietary iron in the therapy of iron deficiency.
The pharmaceutical market is glutted with iron preparations in nearly every conceivable form; each promoted to appeal to physician or patient for one reason or another. The following simple principles may help the physician to find a way through this chaos.
Each dose of an inorganic iron preparation for an adult should contain between 30 and 100 mg of elemental iron. Doses of this magnitude cause unpleasant side effects relatively infrequently.194 Smaller doses have been popular in the past, but these may result in a slower recovery of the patient or no recovery at all.
The iron should be readily released in acidic or neutral gastric juice or duodenal juice (usually pH 5 to 6), because maximal absorption occurs when iron is presented to the duodenal mucosa. Enteric-coated and prolonged-release preparations dissolve slowly in any of these fluids. Thus with such preparations the iron that eventually is released may be presented to a portion of the intestinal mucosa in which absorption is least efficient. Some patients who have been treated unsuccessfully with enteric-coated or prolonged-release iron preparations respond promptly to the administration of non–enteric-coated ferrous salts.
The iron, once released, should be readily absorbed. Iron is absorbed in the ferrous form; consequently, only ferrous salts should be used.
Side effects should be infrequent. This seems not to be a particular problem for any of the common commercially available iron compounds. Despite the claims of pharmaceutical companies, there is no convincing evidence that any one effective preparation is superior in this respect to any other.
Inexpensive iron preparations can be as effective as the more costly ones. The use of preparations containing several therapeutic agents is unnecessary and may increase side effects. Physicians should be aware that if ferrous sulfate is prescribed generically, the choice of preparation is left to the pharmacist who may dispense enteric-coated tablets. It is advisable to specify “nonenteric” or to prescribe by brand name a product that is not enteric-coated. Although substances such as ascorbic acid, succinate, and fructose enhance iron absorption, the gain is offset to a large extent by the increase in frequency of side effects, cost of therapy, or both. There is no convincing evidence to support the use of chelated forms of iron or of iron in combination with wetting agents.
For therapy of iron deficiency in adults, the dosage should be sufficient to provide between 150 and 200 mg elemental iron daily. The iron may be taken orally in three or four doses 1 hour before meals. Infants may be given 6 mg/kg195 daily in divided doses for therapy, or a daily dose of 12.5 mg daily for prophylaxis of iron deficiency
Mild gastrointestinal side effects occur occasionally in the form of nausea, heartburn, constipation, or changes in the stool consistency. A metallic taste may be experienced. The majority of patients tolerate the usual therapeutic doses of iron without the least side effect. However, there is no doubt that some patients, perhaps 10 to 20 percent, experience symptoms that may be ascribed to the iron preparation and may be dose-dependent. In such cases, reduction of the frequency of administration to 1 tablet a day for a few days may alleviate the symptoms; later, the patient may be able to tolerate treatment in full dosage. It might also be useful to change to another iron preparation, especially one with a different external appearance.
Carbonyl iron has been proposed as an alternative to iron salts, on the assertion that it can be given in large doses with minimal side effects. This substance is actually metallic iron powder, with a particle size less than 5 μm. Because it is insoluble, it is not absorbed until converted to the ionic form. The bioavailability of carbonyl iron has been estimated to be approximately 70 percent of that of an equivalent amount of ferrous sulfate.196 Oral doses as high as 600 mg three times daily did not produce toxic effects.196
Widespread iron supplementation in regions where malaria and gastrointestinal infections are highly endemic is associated with increased malaria transmission and childhood mortality, presumably from increased infections.197 Although there is not yet a consensus on optimal strategy in such settings, it seems reasonable to target iron supplementation to children who are iron-deficient.
Acute iron poisoning is usually a consequence of the accidental ingestion by infants or small children of iron-containing medications intended for use by adults. Any potent oral preparation may cause acute iron poisoning, and this serious disorder remains a problem, despite public awareness campaigns and safer packaging of medications.198 In the United States, there were nearly 30,000 reported incidents in 2008. The earliest manifestation of iron poisoning is vomiting, usually within 1 hour of the ingestion. There may be hematemesis or melena. Restlessness, hypotension, tachypnea, and cyanosis may develop soon thereafter, and may be followed within a few hours by coma and death but fatal outcomes are now extremely rare. Usually, medical aid is sought early and, with proper treatment, most iron-poisoned children survive. The initial treatment is prompt evacuation of the stomach. In the home, this may be induced by digital stimulation of the pharyngeal gag reflex. If the patient arrives in the emergency room within minutes of ingestion, gastric intubation and lavage should be performed promptly. Whole-bowel irrigation198 is currently recommended to for all heavy metal intoxications. Supportive measures should be used as needed for shock or for metabolic acidosis should these develop. IV desferrioxamine is the agent of choice for specific therapy of hyperferremia, at a maximum rate of 15 mg/kg per hour for 1 hour, then lowered to 125 mg/hour. Improvement often appears several hours to a few days after onset of iron poisoning. Children who survive for 3 or 4 days usually recover without sequelae. However, gastric strictures and fibrosis or intestinal stenosis may occur as late complications.
As parenteral iron preparations have become safer and easier to administer, the use of parenteral iron is increasing. Established indications for the use of parenteral rather than oral iron include malabsorption, either because of systemic inflammation or gastrointestinal pathology, intolerance to iron taken orally, iron need in excess of an amount that can be absorbed in the intestine, and noncompliance. Parenteral iron administration has an erythropoietin-sparing effect in anemic patients on long-term hemodialysis for chronic renal disease.35,199,200 Because of systemic inflammation and possibly other factors, these patients do not appear to respond adequately to oral iron therapy.
The amount of iron that needs to be given is readily estimated by noting that 1 mL of red cells contains approximately 1 mg of iron. However, various formulas have been used for estimating total dose required for treatment. Because total blood volume is approximately 65 mL/kg and the iron content of hemoglobin is 0.34 percent by weight, the simplest formula for estimating the total dose required for correction of anemia only is as follows:
Assuming normal mean hemoglobin concentration of 16 g/dL, a male weighing 170 pounds, whose hemoglobin concentration is 7 g/dL, would require 170 × (16 − 7) = 1530 mg iron to correct this anemia. To this should be added a sufficient quantity of iron to replete iron stores, approximately 1000 mg for men and approximately 600 mg for women. Thus a 170-pound male with a hemoglobin concentration of 7 g/dL should receive 2530 mg iron.
Parenteral Iron Preparations
Because iron salts are highly toxic when given parenterally, all iron preparations consist of colloidal (nanoparticulate) complex of iron with carbohydrates. To make the iron bioavailable for erythropoiesis and other biologic processes, the iron complexes must be ingested by macrophages and digested so that the administered iron can be gradually delivered to plasma transferrin. Currently available preparations include iron sucrose, low-molecular-weight iron dextran, ferric gluconate, ferumoxytol, ferric carboxymaltose, and iron isomaltoside. High-molecular-weight dextran was associated with anaphylactoid adverse events compared to the other preparations and should therefore be avoided.201 The remaining preparations are safe and serious adverse events are extremely rare.202 Although the recommended methods of administration, including the use of test dosing, the amount of iron per infusion, and infusion rates differ among the preparations, these are not based on comparative studies. Currently available data indicate that the preparations do not significantly differ in safety or efficacy. Premedication to prevent allergic responses was commonly used with the older preparations but it is neither needed nor known to be effective with the newer formulations, and may introduce side effects of its own.
If therapy is adequate, the correction of iron-deficiency anemia is usually gratifying. Symptoms such as headache, fatigue, pica, paresthesia, and burning sensation of the oropharyngeal mucosa may abate within a few days. In the blood, the reticulocyte count begins to increase after a few days, usually reaches a maximum at approximately 7 to 12 days, and thereafter decreases. When anemia is mild, little or no reticulocytosis may be observed. Little change in hemoglobin concentration or hematocrit value is to be expected for the first 2 weeks, but then the anemia is corrected rapidly. The hemoglobin concentration in the blood may be halfway back to normal after 4 to 5 weeks of therapy. By the end of 2 months of therapy, and often much sooner, the hemoglobin concentration should have reached a normal level.
When iron-deficiency anemia does not resolve with oral iron treatment, careful inquiry into the nature, duration, and regularity of iron therapy may reveal a reason for the failure of therapy and permit a gratifying response to be elicited with adequate therapy. Other questions that should be asked in evaluation of such a case are these: (1) Has bleeding been controlled? (2) Has the patient been on iron therapy long enough to show a response? (3) Has the dose of iron been adequate? (4) Are there other factors—inflammatory disease, neoplastic disease, hepatic or renal disease, prior gastrointestinal surgery, concomitant deficiencies (vitamin B12, folic acid, thyroid)—that might retard response? Prominent among these are H. pylori infection, autoimmune gastritis, and celiac disease.15 (5) Is the diagnosis of iron deficiency correct?
Intravenous iron should be effective in patients with established iron deficiency who fail to respond to oral iron after several weeks. Continued loss of blood or, rarely, the genetic disorder iron-refractory iron-deficiency anemia,15 may account for incomplete response to IV iron.
When the cause of the iron deficiency is a benign disorder, the prognosis is excellent, provided bleeding is controlled or can be compensated for by continual iron therapy. If there is a benign cause of recurrent bleeding that is corrected, such as hiatal hernia, menorrhagia, or hereditary hemorrhagic telangiectasia, oral iron therapy may be continued indefinitely; if the bleeding is especially brisk, supplementation with parenterally administered iron or, rarely, with transfusion may be needed. Continuous iron administration may also be required in patients with iron deficiency secondary to intravascular hemolysis with hemoglobinuria.