Chapter 16
Anemia Associated with Pregnancy
Kenneth G. Perry Jr and John C. Morrison
Main Menu   Table Of Contents


Kenneth G. Perry, Jr, MD
Assistant Professor of Obstetrics and Gynecology, Department of Obstetrics and Gynecology, University of Mississippi Medical Center, Jackson, Mississippi (Vol 3, Chap 16)

John C. Morrison, MD
Professor of Obstetrics, Gynecology, and Pediatrics, and Vice Chairman and Director of Research, Department of Obstetrics and Gynecology, University of Mississippi Medical Center, Jackson, Mississippi (Vol 3, Chap 16)



Anemia is one of the most frequent maternal complications diagnosed during pregnancy, occurring in excess of 55% of pregnancies in the United States. It may range from a clinically insignificant laboratory finding to a serious disorder with dire consequences for both mother and fetus or newborn. The word anemia connotes a decrease in the oxygen-carrying capacity of the blood and is best characterized by a reduction in hemoglobin concentration, which may be relative or absolute. Relative anemia is not a true reduction in the red cell mass. The most common example is the apparent reduction in hemoglobin content and packed-cell volume (PCV) resulting from an increase in plasma volume (PV) in the midtrimester of normal pregnancy in the iron-sufficient woman.1 This anemia is a transient phenomenon and is a physiologic event that occurs during pregnancy. Absolute anemia involves a true decrease in red cell mass, is of hematologic importance, and involves increased red blood cell (RBC) destruction, increased corpuscular loss, or decreased erythrocyte production. Another form of classification involves morphologic criteria evident in the anemic process. RBC indices have also been used to emphasize the importance of direct observation of erythrocytes. This classification emphasizes cellular size (microcytic, macrocytic, or normocytic) and staining of the erythrocytes (hypochromic, hyperchromic, or normochromic) and may be helpful in diagnosing the more common types of anemia. Finally, anemias may be categorized according to etiologic pathophysiology for the specific disease process. A combination of these approaches for classification is presented in the following outline:

  1. Relative
    1. Pregnancy
    2. Macroglobulinemia

  2. Absolute
    1. Decreased erythrocyte production (hypoproliferative)
      1. Disturbance in maturation
        1. Hemoglobin synthesis
          1. Heme (iron deficiency)
          2. Globin (thalassemia)

        2. DNA synthesis (megaloblastic)
          1. Folic acid deficiency
          2. Vitamin B12 deficiency

      2. Bone marrow failure
      3. Unknown (neoplasia, inflammation)

    2. Increased erythrocyte loss
      1. Early gestation (abortion)
      2. Late gestation (placental abnormalities)
      3. Puerperium (uterine atony)
      4. Intestinal lesions (hemorrhoids)
      5. Parasites (hookworms)

    3. Increased erythrocyte destruction (proliferative)
      1. Extrinsic mechanism (acquired)
        1. Hypersplenism
        2. Mechanical (microangiopathic)
        3. Coombs-positive anemia

      2. Intrinsic mechanism (inherited)
        1. Membrane abnormalities (spherocytosis)
        2. Enzyme abnormalities (G6PD)
        3. Globin abnormalities (sickle hemoglobin)

      3. Unknown (paroxysmal nocturnal hemoglobinuria)

In the diagnosis of anemia, it is important to remember that there may be more than one etiologic factor involved. Anemia due to folio acid deficiency (FAD), that due to chronic infection, and iron deficiency anemia (IDA) yield different RBC indices in their pure form; if combined, however, a mixed pattern results, which may render these diagnostic tests less useful. Several deficiencies may cause the same type of anemia. Nutritional deprivation, reduced interpregnancy interval, and adolescence are common etiologic factors in IDA in pregnancy. Treatment of one cause may not result in total compensation. Finally, pregnancy itself may alter several common laboratory tests, complicating anemia assessment techniques. During antepartum evaluation of anemia, each laboratory result should be interpreted carefully, since many factors affect the diagnostic and therapeutic modalities used, and thus maternal-fetal outcome.

Back to Top

Although pregnancy is the most common cause of relative anemia, other factors such as splenomegaly, dietary deficiencies (protein), and macroglobulinemia can increase PV, creating an apparent reduction in the RBC count. During pregnancy, most women demonstrate a transient reduction in the RBC mass, hemoglobin concentration, and PCV. This relative anemia is principally due to the larger and earlier increase in PV relative to RBC mass. In most iron sufficient women, the RBC production approaches the extent of increase by the PV at term. However, in IDA or other disorders, the anemia may persist because the production of cells is inhibited and the PV continues to predominate.

Back to Top

The PV begins to rise slowly above the nonpregnant levels at the sixth week of pregnancy (Fig. 1). At 16 weeks, it is approximately 10% above normal; it rises rapidly until 26 weeks to levels greater than 50% of the normal PV in the nonpregnant female and then maintains a constant plateau until near term. Measurement of the PV with the patient in the supine position has shown that the apparent decrease from 28 to 40 weeks' gestation is fallacious. Earlier conflicting observations possibly resulted from a sequestration of fluid in the lower extremities due to caval compression by the enlarged uterus when the patient was in the supine position. Indeed, the transient increase in PV observed at the time of delivery may represent the return of blood from the lower extremities.

Fig. 1. Changes in blood volume (in milliliters) during antepartum, intrapartum, and postpartum portions of human pregnancy.

The mechanism of increase in PV appears to be hormonal, resulting in a retention of sodium and water as pregnancy progresses.1 Increased levels of estrogen, aldosterone, cortisol, prolactin, and human placental lactogen (HPL), which are known to cause sodium retention, are present as early as the sixth week of pregnancy. In addition, the control of antidiuretic hormone appears to be altered during pregnancy, resulting in a positive water balance relative to sodium. Although production of some of these hormones wanes in late pregnancy, the increasing sensitivity of the maternal vascular system to certain agents, particularly aldosterone and prolactin, may continue the hydremia of pregnancy until term.

In the early postpartum period, PV decreases, only to increase again between 2 and 5 days after delivery (see Fig. 1). The increase may be related to the rise in aldosterone observed at approximately 72 hours post partum. This elevation usually abates by 6 days post partum, and the PV begins its final decrease to normal levels. By 3 weeks post partum, the PV is still elevated 10% to 15% above nonpregnant levels, but when measured at 6 weeks, it has usually returned to normal. Although there are many techniques to quantitate the PV changes, dye dilution is the most common. Regardless of the method, it is best to quantitate the PV comparatively in milliliters per kilogram of nonpregnant weight because of the large volume of fluid sequestrations in extravascular spaces and the tremendous intrapatient variations.

There are several advantages to the physiologic increase in PV during pregnancy. The apparent decrease in PCV reduces viscosity, decreasing the possibility of thrombosis. Lowering of peripheral resistance and expansion of the blood volume appear to promote more oxygen exchange at the tissue level for a given cardiac output, reducing cardiovascular work. Also, there is moderate impairment of venous return late in pregnancy, and, since arteriolar sympathetic sensitivity is reduced in the normal pregnancy, an increase in PV may prevent syncope during positional change. Finally, it may afford protection against acute blood loss, allow a graduated response to the hemoconcentration characteristic of such maternal disorders as pregnancy-induced hypertension, and enable the parturient a margin of tolerance for various anesthetic procedures. The only disadvantage appears to be related to the timing of maximal PV increase (i.e., during the second trimester and at delivery) in patients with cardiac or renal disorders who may experience decompensation if their clinical disease is severe.

Routine laboratory assessments can be incorrectly interpreted if the physiologic change in PV is not recognized. The PCV and hemoglobin content may appear to be decreased and, in the opinion of some authors, may lead to unwarranted iron replacement.2 Combined with other processes (e.g., IDA), it may rarely lead to unnecessary transfusions when correction with iron supplementation would suffice. Simple hydremia rarely reduces the hemoglobin level below 10 g/dL. On the other hand; an inadequate expansion of PV may actually mask a failure of RBC mass to increase, resulting in an erroneously high hemoglobin concentration and PCV as found in dehydration, pregnancy induced hypertension, and certain nutritional deficiencies. Other assessments that may be affected by the hydremia include plasma albumin, serum iron, and certain cations or anions (sodium, potassium, chloride, magnesium, and calcium).

The PV, PCV, and total blood volume (TBV) are significantly elevated in parturients with multiple pregnancies and macrosomia. In general, the greater the hydremia, the larger the increase in RBC mass; thus, the initial disparity is balanced in almost all gestations, with severe anemia being a rare corollary. It should also be realized that there is marked individual variation in the amount of pregnancy-induced PV expansion, although there is a tendency for each woman to have the same degree of hydremia in successive pregnancies.

Back to Top

The increase in RBC mass does not begin until approximately 20 weeks, but then increases more rapidly than the PV until 28 weeks (see Fig. 1). From 28 weeks to term, the RBC mass rises only slightly, but the slope of erythrocyte increase begins to exceed that of the PV (a situation opposite that found earlier in pregnancy). The RBC mass is approximately 30% higher than its maximum in the nonpregnant state. In the early postpartum period, the RBC mass remains approximately 10% above nonpregnant levels for 1 to 2 weeks but returns to normal by 6 weeks. The decrease is principally related to blood loss at delivery and a decline in erythrocyte production. There is no evidence of increased RBC destruction during the puerperium. Bone marrow erythropoiesis assumes a normal level of RBC production by the end of the postpartum period.

The increase of RBC mass during pregnancy is accomplished by a complex interaction of several hormonal and physiologic factors, but in general it follows the erythropoietin production. In normal pregnancy, the erythropoietin level begins to rise slowly at 15 weeks, but the effects of this stimulation on RBC mass are first documented at 18 to 20 weeks. The maximal activity for erythropoietin occurs between 20 and 29 weeks, corresponding with the maximal increase in uterine blood flow and basal oxygen consumption. The level of erythropoietin begins to decrease slowly after birth in spite of blood loss at delivery. Studies in mice, but not humans, have shown that the RBC mass is increased in lactating animals compared with non lactating controls. Thus, in some species at least, prolactin may continue its erythropoietic activity in the puerperium. Increased erythropoietin levels are also prominent during hypoxia, phlebotomy, polycythemia, some anemias (those due to iron, vitamin B12, or FAD but not those due to starvation, infection, most malignancies, and chronic renal disease), and hypernephroma or other erythropoietin-secreting tumors. It is decreased by hyperoxygenation, ordinary transfusion, uremia, and malnutrition.

Erythropoietin production and the subsequent size of the erythrocyte mass are directly related to increased basal oxygen consumption, an event associated with pregnancy. Other factors, such as elevation of cardiac output, decrease in peripheral resistance, reduction in viscosity, and increased erythrocyte content of 2, 3-diphosphoglycerate, are also related to the increased need of maternal and fetal tissues for oxygenation. This oxygen requirement stimulates the kidneys as well as other organs to elaborate renal erythropoietic factor. This precursor transforms a dormant circulating prehormone into erythropoietin, a glycoprotein with a molecular weight of 60,000 to 70,000 that is found in the plasma and urine. It stimulates the genetically predetermined precursor stem cells in the bone marrow to differentiate by way of the erythroid cell line into erythrocytes (Figure 2). Prolactin also appears to enhance the effect of the erythropoietin already produced. HPL, by its general anabolic action, may support endogenous erythropoietin. The site of HPL action appears to be at the level of the stem cell, whereas prolactin seems to act on the late erythroid precursors.

Fig. 2. Time sequence of the maturation process of erythrocytic cell lines (erythropoiesis), revealing metabolic determinants and cellular appearance at each stage from immature stem cell in the marrow to destruction of mature red blood cells of the reticuloendothelial system.

Therefore, during normal pregnancy, the PV rises early and creates a relative or transient anemia during the first and second trimester. The RBC mass rises to its maximal level during the third trimester, and in iron-sufficient women the difference in PCV and the concentration at term is minimal. The TBV, as shown in Fig. 1, is composed of and follows the increase in PV and erythrocytes.1 The significance of all the changes previously discussed are related to the diagnostic and therapeutic considerations of the specific anemias later in the chapter. It must be remembered, however, that failure to take these physiologic changes into account may confuse the clinician in cases of absolute anemias.

Back to Top

As in most diseases, the diagnosis of anemia rests on a thorough history, physical examination, and laboratory assessment.3 Most of the symptomatology and physical signs of anemia can be attributed to a reduction in the oxygen-carrying capacity of the blood. Tissue hypoxia usually does not occur in the patient with anemia except in the most severe cases, although relative hypoxia can be observed as a result of increased oxygen consumption during pregnancy. In most cases, however, a compensatory expansion of RBC mass occurs to offset this process. The decrease in peripheral resistance, modest elevation of cardiac output, increased tissue perfusion, and hypocapnia seen during pregnancy may ameliorate anemia during pregnancy. Obviously, precursors such as iron, vitamins, and folic acid must be present or the anemia may proceed uncompensated, It should be remembered that severe, primary hematologic disease usually occurs infrequently during the childbearing years; however, hematologic manifestations secondary to other diseases occur as frequently in the pregnant patient as in the non gravid patient.


The history should include general symptoms (e.g., evaluation of the performance status of the patient), which may be helpful in establishing the magnitude of the anemia as well as in delineating the effect of therapy. Symptoms of mild anemia, such as easy fatigability and malaise, are common in many normal pregnancies. Most of the patients with anemia during pregnancy of mild to moderate degree have no additional manifestations. Patients presenting with the classic symptoms of tachycardia, exertional dyspnea, pallor, and palpitations should be carefully evaluated. Not only are these manifestations of moderate to severe anemia but these symptoms may also herald a rare underlying hematologic disorder such as leukemia or cardiorespiratory disease. Adolescent pregnancy, frequent successive pregnancies, nutritional deficiencies, and concurrent medical diseases may contribute to an anemia process. A notation of reaction or exposure to various drugs and chemicals may be an important factor in diagnosis of hemolytic anemias. Finally, a family history may be helpful, particularly as it relates to hemoglobinopathies or inherited hematologic disorders.

Physical Examination

On the physical examination, a central feature of anemia appears to be the pallor caused by the reduced hemoglobin level. This is most helpful in whites and Eurasians, but examination of the mucous membranes may be used for similar purposes in blacks. During pregnancy, however, examination of the skin and mucous membranes in any race may be misleading owing to the hyperemia of these areas. Palmar creases, which usually are white in the fully open hand if the hemoglobin level is less than 10 g/dL, may appear pink in anemic pregnant patients because of hyperemic effects of HPL and progesterone. Pallor of the nail beds, however, is a reliable indicator of anemia during pregnancy in any racial group. Nails that are ridged longitudinally and flattened (koilonychia) rather than convex are present in chronic IDA. Other observations that may be helpful include cyanosis (congenital methemoglobinemia) and jaundice (hemoglobinopathies and hemolytic processes). An enlarged, smooth tongue is associated with pernicious anemia, but is quite rare in this country. Glossitis related to IDA is more common; the tongue is coated, enlarged, and painful. With severe IDA, the lips may reveal cracks, particularly at the edges (cheilosis). Neurologic examination may be needed if vitamin B12 deficiency or IDA is suspected, since both may give rise to peripheral neuropathies. These signs are not altered by pregnancy but, except for pallor, are usually not present unless the anemia is severe (hemoglobin <6 to 8 g/dL). It is also important that the skin, liver, spleen, and lymph nodes be evaluated for enlargement, excoriation, or other abnormalities that may indicate primary hematologic disease or a secondary response to other disease states.

Laboratory Examination

Since anemias are so common in women of reproductive age and since most women with mild to moderate anemias during pregnancy are asymptomatic, it is recommended that all patients be assessed for anemia during their initial prenatal visit. Laboratory assessment includes a complete blood count (hematocrit, hemoglobin concentration, white blood cell count, and platelet count), as well as a peripheral blood smear. The laboratory should examine carefully the RBC morphology in all prenatal patients, because it may reveal several hematologic abnormalities. Because there is a demonstrated association between maternal anemia and unfavorable pregnancy outcome, these assessment tests for anemia appear to be cost effective and beneficial for the parturient and fetus. Normal screening and specific values for the pregnant and the nonpregnant state are shown in Table 1.

TABLE 1. Laboratory Norms for the Nonpregnant and Pregnant Patient




General Assessment


12.5–14 g/dL

11.5–12.5 g/dL

 Packed cell volume



 RBC count

4.2–5.4 million/mm3

3.8–4.4 million /mm3

 Mean corpuscular volume

80 cu μ100 cu μ/cell

70 cu μ–90 cu μ/cell

 Mean corpuscular hemoglobin

27–34 pg /cell

23–31 pg/cell

 Mean corpuscular hemoglobin concentration

31–36 g/dL


 Reticulocyte count



Specific Diagnostic Tests

 Serum iron

50–110 μg/dL

40–100 μg/dL

 Unsaturated iron binding capacity

250–300 μg/dL

280–400 μg/dL

 Transferrin saturation



 Serum ferritin

75–100 μg/L

55–70 μg/L

 Free erythrocyte protoporphyrin

25 μg/L

35 μg/L

 Estimated sedimentation rate

0–15 mm/hr

40–50 mm/hr

 Serum folate (fasting)

6.5–19.6 ng/ml

5–10 ng/ml

 Serum B12

150–450 pg/ml


The laboratory assessment of anemia is more difficult during pregnancy. In general, iron-sufficient, disease-free women with relative anemia during pregnancy have a hemoglobin level above 11 g/dL and a PCV (or hematocrit) above 35%. The average hemoglobin levels during pregnancy are between 11.5 and 12.5 g/dL instead of the normal 13 g to 15 g/dL found in the nonpregnant state (see Table 1). The hemoglobin content during pregnancy tends to be further reduced, since the derivative actually measured by most laboratory techniques (cyanmethemoglobin) is slightly lower during gestation. The PCV or hematocrit likewise is lower in pregnancy, averaging 33% to 38% compared with the 37% to 47% range associated with the normal nonpregnant female. Since the relationship of hemoglobin level to PCV is more erratic in pregnancy, the measurements of RBC mass and other indices (i.e., mean corpuscular volume, MCV; mean corpuscular hemoglobin, MCH; and mean corpuscular hemoglobin concentration, MCHC) may not be as helpful as during the nonpregnant state. Of these, the MCV appears to be a good discriminator of the various types of hypoproliferative anemias (Fig. 3). All indices reflect average cell values and do not detect abnormalities in mixed cell populations. A patient with an IDA and concomitant megaloblastic process may reveal normal indices rather than the classic presentation of either disease.

Fig. 3. Anemias and related laboratory findings.

If the initial assessment of the patient's hematologic state indicates an anemia, the classification by mechanism as outlined in Fig. 3 can allow the physician to arrive at a diagnosis by an orderly method of laboratory analysis. The reticulocyte count (RC) divides anemias into categories in which the marrow is proliferative or hypoproliferative. Methylene blue is used to stain the reticulocytes (young, nonnucleated RBCs) on the peripheral smear. Usually if the RC in the presence of anemia is below 3%, the mechanism is one of diminished erythropoiesis. If it is above 3%, excessive hemolysis or acute blood loss is usually the cause. Therefore, a normal RC of 1% to 2% during pregnancy in a patient with anemia indicates a hypoproliferative process in which the patient cannot respond with new RBC production. Examination of the bone marrow is usually not performed during pregnancy owing to hypervascularity and subsequent maternal risk. However, if serious hematologic disease is suspected, then this diagnostic technique is usually required. Specific laboratory methods used during the diagnosis of anemia are discussed under the particular anemia processes. A flow sheet as shown in Fig. 3 allows the orderly procurement of essential laboratory tests in the diagnosis of anemia.

Back to Top

Maternal Considerations

Since severe reductions in hemoglobin levels are rare with the transient anemia of pregnancy, women with a hemoglobin level below 10 g/dL should be considered to have an absolute anemia. Serious maternal consequences directly related to the anemia are rare in women with levels between 6 and 10 g/dL, although the process may lead to serious morbidity (Fig. 4). Indirectly, IDA may be related to obstetric complications such as postpartum hemorrhage, operative delivery, and placental abnormalities (abruptio placentae or placenta pre via).3 Morbidity such as infections, prolonged hospital stays, and general recuperative health problems have also been associated with preexistent anemias, although correlation with nutrition and socioeconomic status cannot be ignored. In contrast, patients with hemoglobin levels below 4 to 6 g/dL may face life threatening problems owing to high output congestive heart failure with decreased oxygenation of cardiac tissue.

Fig. 4. Percentage of pregnant patients with various symptoms, complications, or disease processes who have in common an anemia (packed-cell volume < 30%).(Modified form Kitay DZ: Assessing anemia in the pregnant patient. Contemp Ob Gyn 2:17, 1973)

Placental Considerations

The effect of severe anemia on the placenta is less well defined. If the hemoglobin level is 6 to 10 g/dL, there is usually placental hypertrophy. Although this enlargement may be beneficial, the placental size has not been correlated with improved maternal or fetal outcome. Anemia is commonly associated with other maternal disease processes. Pica and pregnancy-induced hypertension are associated with IDA; chronic renal disease and abruptio placentae with FAD; and labor dystocia with sickle hemoglobinopathies. Some of these correlates remains unsubstantiated and may not be causally related.

Fetal Considerations

The effect of maternal anemia on the fetus is less clear. Anemia, regardless of etiology, has been associated with prematurity, low-birth-weight infants, abortions, and fetal deaths, even when the process is mild (hemoglobin levels of 8 to 11 g/dL). Although reported in individual cases, a direct relationship between anemia and fetal distress has not been established unless the anemia is severe (hemoglobin levels <6 g/dL). Also, mild deficiency states (FAD, even without anemia) have been related to decreased birth weight or fetal loss. However, these findings cannot be substantiated as cause-and-effect correlates, since the effects of nutrition and social status cannot be separated. On the other hand, most would agree that severe hemoglobinopathies are associated with a higher incidence of premature and low-birth-weight infants, as well as with increased fetal loss.

In assessing a direct causal relationship between anemia and adverse maternal-fetal effects, one must remember that the association may be a reflection of many factors other than anemia. The common association and interaction of anemia, poor nutrition, and lower educational norms among the socioeconomically disadvantaged with poor maternal-fetal outcome is well known. The influence of factors such as smoking, poor obstetric history, and inadequate prenatal care must receive further assessment before a direct association of an anemic process and pregnancy outcome can be made. Therefore, it appears that while severe anemia from any cause may have a direct adverse effect on the parturient and fetus or newborn, the effect of mild to moderate anemia is less clear but may be significant.

Back to Top

Disturbance in Maturation: Hemoglobin Synthesis


IDA accounts for 75% of all the anemias diagnosed during pregnancy.4 Iron depletion without signs of anemia is very common during pregnancy. In one study, iron stores were found to be exhausted in 25% of young, apparently healthy women on their first prenatal visit. Another study demonstrated that 80% of normal pregnant, iron-sufficient women if not given iron supplementation were anemic by term. Finally, it has been estimated that 20 million persons in the United States are iron depleted, particularly in the socioeconomically disadvantaged population. Any discussion of iron deficiency necessarily includes an understanding of iron absorption, metabolism, and loss.5 Iron is necessary for RBC formation but is also intimately involved with proteins important in intermediary metabolism. Nearly half the enzymes and cofactors in the Krebs cycle either contain iron or require its presence.

Iron Compartments.

The basic compartments of iron distribution include hemoglobin iron, storage iron, myoglobin iron, labile pool iron, other tissue iron, and transport iron. Table 2 illustrates the significant differences in the total body iron distribution between the adult male and the female during the reproductive years.6 Pregnancy normally increases the amount of iron but not the percentage distribution to each compartment; however, socioeconomic status, nutritional status, and concurrent disease processes can modify the dispersion of iron in each compartment.

TABLE 2. Iron Compartments*

Iron Content (mg)




Total Body Iron (%)

Hemoglobin iron




Storage iron




Myoglobin iron




Labile pool




Other tissue iron




Transport iron




* These values represent estimates for an iron-sufficient person.
† Nonpregnant.
Hemoglobin Iron.

Assessments with radioactive tracers have shown that hemoglobin iron makes up 65% to 70% of total body iron and averages 1700 mg in a normal adult woman. During pregnancy, hemoglobin iron is increased by 20%. Hemoglobin contains 0.34% of iron by weight; thus, 1 ml of packed red cells contain approximately 1 mg of iron. Therefore, iron loss or gain can be calculated if the change in RBC mass is known. It is principally the hemoglobin iron that is reflected in the laboratory test such as PCV or hemoglobin concentration. Initially, the storage iron compartment must be depleted before clinical signs or a change in laboratory assessment reveals actual iron deficiency in hemoglobin content.

Storage Iron.

Storage iron exists in two forms: ferritin and hemosiderin. In healthy, iron-sufficient women, it totals approximately 600 to 800 mg and makes up approximately 27% to 30% of total body iron. Depletion of this storage compartment occurs when iron loss or use exceeds iron absorption; it is usually decreased during pregnancy even in iron sufficient women.

Ferritin contains approximately half the storage iron and is found in plasma as well as in most of the cells in the body; several varieties of ferritin have been demonstrated by electrophoresis and isoelectric focusing. Ferritins indigenous to many organs and to reticuloendothelial cells have been identified, supporting the hypothesis that these iron proteins are organ specific and are products of different genotypes. Ferritin has been identified by electron microscopy as a water-soluble complex with a molecular weight of approximately 460,000, which is 20% iron. Stereochemically, it exists as a complex of ferric hydroxide and the protein carrier apoferritin. Each apoferritin shell houses a core of ferric hydroxide and phosphate ions, which are dispersed in a rigid latticelike arrangement. The role of ferritin in iron absorption is direct; a low concentration of this compound in the intestinal mucosal cell enhances the biosynthesis of additional apoferritin, which results in more iron absorption. The life span of apoferritin is only a few days; the degeneration and resynthesis provide an available intracellular iron pool. The measurement of serum ferritin has become one method to delineate iron stores without having to resort to bone marrow sampling.

Hemosiderin is found only in cells of the reticuloendothelial system, such as those of bone marrow, liver, and spleen (about one third in each organ). Immunologic studies have shown that the protein components of both ferritin and hemosiderin are identical; hemosiderin may represent a partially denatured ferritin. Hemosiderin is approximately 25% to 30% iron by weight, is water soluble, and, in contrast to ferritin, can be seen with the light microscope. It represents about 50% of the storage iron and about 12% to 15% of the total body iron. Hemosiderin represents that quantity of iron demonstrated by Prussian blue stains of bone marrow aspirates. This storage iron is attached to a substrate called apohemosiderin, which is amorphous, thus lacking the firm, crystal like molecular arrangement of ferritin. The mechanism of mobilization, use, and contribution of storage iron in an anemia process is discussed in the sections on iron absorption and homeostasis. While the measurement of ferritin is involved with an assessment of iron absorption capabilities, the assessment of hemosiderin is a measure of iron balance; until hemosiderin becomes depleted, no signs of iron deficiency develop (Fig. 5). Therefore, as long as storage iron is present and released normally, no change in the peripheral blood smear is seen.

Fig. 5. Relationships of known iron compartments to various stages of clinical iron depletions as depicted in the cellular appearance of the RBC.

During pregnancy, the increase in RBC mass gives an indirect measurement of iron stores. The principal cause for a failure of the expansion of the PCV is the absence of bone marrow hemosiderin. This finding indicates exhaustion of storage iron, its absence being the earliest sign of iron deficiency. If IDA is clinically evident by a decreased hemoglobin concentration; by a characteristic microcytic, hypochromic blood film; and by altered RBC indices, the iron stores will be nonexistent. On the other hand, bone marrow stores decrease to minimal levels during pregnancy, even in iron sufficient women, in the absence of iron supplementation. Since a bone marrow examination is not practical during pregnancy unless severe hematologic disease is suspected, reductions in serum ferritin, PCV, hemoglobin concentration, and RBC indices are the first laboratory indications of inadequate iron stores. Iron stores are reduced in IDA, blood loss, and nutritional anemias; however, hemolytic processes, hereditary anemias, and ineffective iron utilization during infection or inflammation may be associated with normal to increased storage iron levels and concomitant reductions in hemoglobin levels, PCV, and RBC indices.

Myoglobin Iron.

Myoglobin iron accounts for approximately 3% to 4% of total body iron (130 mg) and is relatively constant between men and women, even in pregnancy. Each myoglobin molecule consists of a heme moiety surrounded by a helical chain of 150 amino acids. The myoglobin is less than 1% iron and has a molecular weight of approximately 17,000. It is found in most muscle cells and appears to serve as an oxygen reservoir when cellular damage from hypoxia occurs. Function of this iron component peculiar to pregnancy has not been reported.

Labile Iron Pool.

The labile iron pool represents approximately 80 mg of iron constantly in exchange between the plasma, interstitial, and intracellular compartments. An intracellular protein produced by the liver appears to be responsible for short-term binding and release of the labile iron. This protein has been termed acetate-extractable ferroprotein and has a molecular weight of approximately 12,000. It has been isolated from cells in the lung, liver, intestinal mucosa, spleen, and kidneys, as well as from RBCs. Iron kinetic studies have shown that approximately 80% of this transitional protein eventually is reincorporated into hemoglobin and that the remaining 20% is associated with storage compounds. The labile iron pool offers a method of studying the clearance rate, incorporation data, and daily iron turnover; ferrokinetic studies indicate that in IDA the incorporation of iron into hemoglobin and subsequently into normal RBCs is almost 100%. In contrast, in hemolytic or megaloblastic anemia, the incorporation of iron is rapid and complete, but there is ineffective erythropoiesis and early destruction of the erythroblasts. Patients with aplastic anemia or thalassemia also demonstrate an incorporation rate that is markedly reduced in the presence of normal erythrocytes.

Enzymatic Iron.

Tissue iron is constant during pregnancy and represents approximately 6 to 8 mg of iron, or between 0.2% and 0.5% of total body iron. Although extremely small, this compartment is very important because it includes enzymes such as the cytochrome peroxidase catalase dehydrogenase, as well as acyl-CoA and various other oxidases. Changes in enzymatic iron are reflected in decreased efficiency of the mitochondrial systems; thus, severe anemias may affect fetal growth and development by such a mechanism. In addition, maternal-fetal steroidogenesis and host response to various disease states may be modified by functional changes in these critical enzyme systems. As shown in Fig. 5, the amount of parenchymal iron is altered only in severe IDA, and it is one of the last compartments to become depleted.

Transport Iron.

Transport iron represents less than 0.1% of the total iron (3 mg) in both men and women, and yet, kinetically, it is the most active compartment, being replaced approximately every 2.5 hours. The metal in this category represents the interchange between all compartments previously mentioned and is not altered during pregnancy. Transport iron is loosely bound to a specific protein (i.e., transferrin), as shown in Fig. 6. Transferrin migrates electrophoretically with the β-globulins and has a molecular weight of approximately 80,000; it is called apotransferrin when not bound to iron. Although the kinetic properties of each apotransferrin are similar, cellular specificity is evident, since 19 genetically distinct molecular variants have been described. It is synthesized principally by the liver but also by other tissues of the reticuloendothelial system and is the only iron form that shows diurnal variation; the highest values are obtained in the morning.

Fig. 6. Transportation sequence of iron involving intestinal absorption, marrow hemoglobin incorporation, RBC destruction, return to labile iron pool, and storage deposition.

A trivalent iron atom may be bound at both ends of the polypeptide chain (see Fig. 6), and binding affinity is increased when one site is occupied. Transferrin receives absorbed iron and transports it to the immature marrow normoblasts. The transferrin in the plasma in iron-replete subjects is one third saturated, leaving two thirds unoccupied; this is measured as the unsaturated iron binding capacity (UIBC). When the serum iron is decreased, the UIBC is increased and vice versa.

Iron Absorption.

Iron absorption depends on many factors (Table 3), reflecting the diet, the status of the bowel lumen, and mucosal abnormalities, as well as systemic factors.7 Iron is principally absorbed in the duodenum and proximal small intestine. The iron presented to the gastrointestinal tract is usually in one of three forms: the ferrous form (from elemental iron), hemoglobin (from animal protein sources), and the trivalent or ferric form (from vegetable complexes) (Fig. 7). The ferrous salts are best, since they need no conversion to be absorbed; the ferric iron in vegetable protein complexes must be reduced to the divalent state before it can pass into the mucosal cell. Hemoglobin iron is readily absorbed after being hydrolyzed in the gut lumen into heme and globin. The globin, although degraded by intestinal enzymes into small peptides, remains an integral factor in absorption, since it continues to stabilize the heme iron in the ferrous state.

TABLE 3. Conditions Affecting Iron Absorption


Increased Absorption

Decreased Absorption

Iron content

 Form of iron

Heme iron

Enteric-coated capsules


Adequate ferrous salt



Iron deficiency


Intraluminal factors

 Intestinal secretions

Hydrochloric acid






Intrinsic factor


 Stomach contents

Ascorbic and other acids,

Oxalates, phytates, phosphorus,




 Intestinal motility





EDTA, desferrioxamine

 Metallic cations


Cobalt, nickel

Mucosal factors

 Disease states

Intermittent outlet obstruction

Gastrectomy, lymphoma



Chronic diarrhea (sprue)


Decreased mucosal iron

Increased mucosal iron

Systemic factors


Acute blood loss

Aplastic anemia


Hemolytic anemia

Transfusion, chronic infection




 Iron requirements

Pregnancy, growth

Weight loss, thalassemia


Iron deficiency


Fig. 7. Mechanism of iron absorption from the intestinal lumen through the mucosal epithelium and into various iron compartments.

Gastric acid degrades the organic iron complexes (vegetable protein) and reduces the trivalent mineral to ferrous iron. Moreover, hydrochloric acid serves to prevent alkaline dietary constituents from forming insoluble complexes with ferrous iron, to prevent various agents from oxidizing iron to the ferric state, and to prevent iron precipitating as insoluble ferric hydroxide. While gastric juice may facilitate elemental iron absorption, it does not appear to have any effect on hemoglobin iron absorption. The alkaline milieu of the duodenum and upper small bowel favors absorption, since the ferrous iron has greater solubility in these surroundings; this is true for all three sources of iron, but it is particularly necessary for hemoglobin iron. Therefore, agents that form alkaline-soluble complexes promote iron absorption; these include the reducing agents ascorbic acid, fructose succinic acid, lactate, pyruvate, and sorbitol and certain amino acids such as cysteine. Certain drugs such as hydroquinone and alcohol may also increase iron absorption by increasing gastric acid production, while medications that decrease intestinal motility, such as atropine or reserpine, also enhance iron absorption. Bile may have a limited role in facilitating iron absorption, but pancreatic secretions are thought to be noncontributory.

In contrast, oxidants and certain medications reduce iron absorption by increasing the concentration of the trivalent form. Other substances form insoluble complexes with the ferrous iron; chelators have a similar mode of action. Metallic cation competitors decrease absorption by competing for binding sites, while other agents reduce absorption by reducing gastric acidity. Cathartics and disease processes reduce iron absorption by increasing intestinal motility, reducing the time of exposure of mucosal cells to iron. Surgical removal of the gastric or duodenal mucosa limits the surface area for absorption.

Finally, another factor, pica, has long been associated with IDA. Pica is the ingestion of various substances that have no dietary value; pagophagia (ice), geophagia (clay), and amylophagia (starch) are examples. In some ethnic groups, it is a common associated finding in over 60% to 80% of patients with proved IDA. The “cravings” these patients experience are not always for nonfoods, but may be for vegetables (carrots or celery) or commercial items such as “fast foods” (invariably not a good source of iron). Formerly all forms of pica were thought to retard absorption of iron. Now it is clear that pica is a sign of IDA rather than a cause. A history of pica in an adult suggests that the patient has IDA, since the abnormal craving usually disappears in a few days after the institution of iron supplementation. It is not proved but may be inferred that certain food fetishes during pregnancy may represent pica due to the increased iron demands of pregnancy. This hypothesis is supported by the observation that most pica occurs after the 20th week, a time during which the iron demands increase most rapidly. The relationship of pica to IDA is not completely understood; however, it seems to be related to a decrease of iron-containing enzymes in the oral mucosal cells of these patients.

The regulatory mechanism of iron uptake is closely related to the plasma iron turnover rate (see Fig. 6). It appears that transferrin from the labile pool is deposited into the developing mucosal cells destined to form the basal layer of the crypts of Lieberkühn in the duodenum. This protein controls the amount of iron absorbed by trapping the metal entering the cell. The conversion of the absorbed ferrous iron into ferritin is a metabolic process that requires a chelator, such as ascorbic acid, an energy source (probably adenosine triphosphate), and, finally, oxidation of the divalent iron to the ferric iron, which requires ceruloplasmin (see Fig. 7). This sequence is adequate for elemental iron and vegetable protein iron absorbed as the simple ferrous cation. However, hemoglobin or myoglobin iron must be further degraded by a microsomal enzyme called heme oxygenase. Once the mucosal cells are saturated with ferritin iron absorption is halted. If the iron stores are reduced, very little ferritin is present in the mucosal lining and iron is absorbed rapidly through the cell, by transferrin, into the bloodstream. In contrast, during iron sufficiency, increased amounts of cellular ferritin are present and the iron is either stored as ferritin or is not absorbed. Ferritin is lost as the duodenal mucosal cell matures and works its way from the base of the crypt to the luminal border. The mucosal cell control of absorption is not, however, absolute; it can be overcome if the concentration of iron at the intestinal lumen increases sufficiently, as in toxic overdoses. If large amounts of iron are presented chronically, hemochromatosis may eventually result.

The transport of iron from the mucosal cell to the utilization pool is performed by transferrin (see Fig. 6). When iron is required by the marrow, transferrin obtains trivalent iron from the ferritin in the mucosal cells. The ferric iron binds to one of the glycoprotein moieties of transferrin, is transported to the marrow, and is used for hemoglobin synthesis. The transferrin binds briefly on the normoblast membrane, where it delivers one or both iron atoms. During the binding process, transferrin may partially enter the cell by a process of endomitosis and actually participate in hemoglobin formation. The release of iron is an active process that can be abolished by enzyme inhibitors. Although direct transfer of transferrin to the normoblast is thought to be the major pathway, rhophenocytosis (accepting of transferrin by reticuloendothelial cells) may represent a minor pathway. Once the iron transferrin complex is attached to the receptor site inside the normoblast, it is conjugated to a protein molecule with a molecular weight of approximately 20,000 and transported to the mitochondria, where heme synthesis occurs; this process takes 8 to 10 minutes. As the erythrocyte matures, it retains its capacity to take up iron and synthesize heme until after the reticulocyte stage, in which the mitochondria are lost (see Fig. 2). After the erythroid cells mature, they are extruded from the bone marrow into the peripheral circulation and have a normal life span of 120 days. The life cycle is complete when the red cells are destroyed and hemoglobin degradation takes place within the reticuloendothelial system, chiefly the liver and the spleen (see Fig. 6). Digestion of the red cells occurs within a few hours, and the plasma iron is redistributed, 80% into new hemoglobin and 20% as storage iron. The mechanism by which the reticuloendothelial cells transfer the iron to transferrin appears to involve ceruloplasmin (ferroxidase), which is also involved in the binding of iron within the mucosal cells.

Iron Homeostasis.

The conservation of iron in humans is tenacious, with only 0.1% of the total amount of body iron lost each day. This amount is easily replaced in the nonpregnant adult if the dietary source is adequate. The average amount of iron excreted by the adult averages 0.9 mg/day, with most being lost in the intestinal tract as desquamated gastrointestinal cells, blood, and bile. Additionally, epidermal cell loss and sweat produce a dally iron loss of 0.2 mg. In areas of high temperature and humidity, an additional 0.5 mg/day may be lost, but this loss rarely produces IDA. Finally, a small amount (0.1 mg) of iron is excreted dally in the urine. Both sexes lose a similar amount of iron through these mechanisms; the basal loss is approximately 14 μg/kg/day. In women during their reproductive years, the iron losses are compounded by menses. Although the blood loss is relatively constant in successive periods, the individual variation between women is large. Most normal, iron-sufficient women lose an average of 25 to 45 ml of blood through menses each month, which approximates 0.7 to 1.4 mg/day in terms of iron loss. The blood loss is lower in the younger age-groups, in women taking oral contraceptives, in women with good nutrition, and in higher socioeconomic groups. Blood loss is exaggerated in women over 40 years old, those with intrauterine contraceptive devices, patients after tubal sterilization, females of high parity, women of lower socioeconomic groups, and those with poor nutrition. A menstrual blood loss of 50 to 60 ml seems to be the upper limit of normal, since women whose losses have exceeded this amount eventually develop IDA. Therefore, the total basal iron loss of a woman in the reproductive years would appear to average 1 mg for menstrual loss in addition to the 1 mg obligatory loss experienced by men and women alike.

Many factors, both physiologic and those related to disease states, can cause a healthy woman to become iron deficient. Pregnancy has a marked effect on iron homeostasis (Table 4).8 In healthy menstruating women, the loss of 2 mg/day can be overcome by a daily food intake: of 1800 to 2200 calories, which contain 11 to 13 mg of iron. However, even in an iron-sufficient state, large amounts of iron must be borrowed from iron stores to complete a pregnancy (Table 4). During the first half of pregnancy, iron requirements are not increased; in the absence of menses, an intake of 11 to 13 mg/day is adequate. After the 20th week, however, the RBC mass begins to expand and the fetus requires more iron. Even with increased absorption, the amount of dietary iron is not adequate to prevent a reduction in iron stores. Obviously, if dietary iron does not meet the requirements, then the storage iron must supply the needs. Since the average North American diet provides about 6 mg of iron per 1000 calories and the pregnant woman of ideal weight should consume 1800 to 2200 calories each day, the average pregnant woman should receive between 10 and 12 mg of elemental iron daily.9 Even though the absorption rate increases from 10% in the first trimester to 30% during the latter half of pregnancy, the iron acquired from the diet alone during this time ranges from 450 to 500 mg; thus, approximately 400 to 500 mg must be supplied by storage iron during pregnancy. In women who are deficient in storage iron prior to pregnancy, this further requirement may lead to overt IDA. It should also be noted that if the storage iron is insufficient at the beginning of pregnancy, the maternal hemoglobin mass will not be expanded until the fetal demands are met. The lack of expansion of the RBC mass thus may be an indication of inadequate iron stores. Post partum, the amount of iron in the expanded hemoglobin mass not lost at delivery can be returned to the iron stores and the anemia can be partially balanced. However, in most women this is not sufficient replacement, and almost all women will remain deficient in storage iron unless they receive supplementation during gestation.

TABLE 4. Iron Homeostasis in Pregnancy,*Expressed in Milligrams of Elemental Iron

Maternal Loss

Maternal Gain




From Stores

500 (RBC mass

490 (diet)†





300 (fetal/placental)

270 (returned to storage



 after delivery)


190 (basal loss)



230 (loss at






TOTAL: 1220

TOTAL: 760


* Lasts 20 weeks.
† Average of 3 to 4 mg/day actual absorption.

Laboratory Assessment.

The laboratory diagnosis of IDA depends on the severity of iron depletion. In the mildest stage, iron deficiency is manifested by a decrease in serum ferritin, but serum iron, PCV, and hemoglobin values are usually normal (see Fig. 5). Iron deficiency without anemia is the next stage; absence of storage iron, manifested by reduced serum ferritin, low serum iron, and decreased transferrin saturation without anemia are characteristic. IDA is first reflected by a reduced RBC mass, then reductions in PCV and hemoglobin levels following hypochromasia and microcytosis (see Fig. 5).

Iron assessment tests must be altered during pregnancy. A relative decrease in serum iron occurs from approximately 12 weeks through 32 to 34 weeks owing to the increase in PV. However, as the RBC mass approaches the increase in PV, the serum iron rises to normal nonpregnant levels. During the early postpartum period, serum iron again rapidly decreases over the first 4 to 5 days before returning to normal at the end of the first week. This is probably related to ineffective release of storage iron owing to the change in hormonal milieu. Transferrin begins to increase from 12 weeks through 34 to 36 weeks, but a slight decrease occurs toward term. During the first 7 days after delivery, the transferrin concentration increases before it returns to normal levels approximately 10 days post partum. These changes are also thought to be hormonally mediated, since similar changes have been observed in women taking oral contraceptives. These observations are present only in iron sufficient pregnant women and are not significant enough to affect the diagnosis of IDA.

The most consistent findings in the blood of a patient with IDA is a decrease in the PCV and hemoglobin concentration, with concomitant hypochromia and microcytosis observed in the blood film. Serum iron, serum ferritin, UIBC, and transferrin saturation may be used to confirm IDA, although they are not routinely obtained during antepartal screening.10 IDA is usually suspected if the serum iron is below 60 μg/dL, the serum ferritin is below 30μ g/L, the transferrin saturation is below 20%, and the UIBC is above 350 μg/dL; the normal values are shown in Table 1. Values of less than 30 μg/dL, less than 10 μg/L, less than 10%, and greater than 400 μg/dL in the serum iron, ferritin, transferrin saturation, and UIBC, respectively, are diagnostic of IDA. Some authors have advocated further evaluation in an attempt to detect early iron-deficient states; however, these have yielded limited clinical information.

In iron-deficient states, zinc may replace iron in the protoporphyrin ring; thus, the measurement of RBC zinc protoporphyrin (RBC ZP) may be an accurate predictor of IDA. This test is relatively rapid and is not costly, although it is not specific for IDA. In addition, the measurement of iron chelating agents such as desferrioxamine, followed by measurement of mobilized iron stores excreted in the urine, has been of value in some cases. The measurement of iron stores by nuclear magnetic resonance may also be helpful in cases of early IDA. In mild cases of IDA, the measurement of free erythrocyte protoporphyrin (FEP) is increased fivefold. Measurement of FEP has been advocated as a screening test for IDA, but the results are not related to the severity of the anemia, nor is it specific for IDA.

Differential Diagnosis.

Although iron depletion or IDA can be easily treated in most cases, it is important to rule out more severe hematologic or systemic diseases. Hypochromic anemia, caused by the following conditions, may be confused with IDA: (1) chemical toxicity related to the intake of chloramphenicol, lead, alcohol, or isoniazid; (2) inflammatory processes; (3) malignancy; (4) pyridoxine-responsive anemia; and (5) hemoglobinopathies.

Drug effects may be toxic (lead, alcohol) or idiosyncratic (isoniazid). Maternal encephalopathy and basophilic stippling of the normal red cells may be observed with lead poisoning. On the other hand, toxic dosages of alcohol or reactions to drugs (chloramphenicol) usually cause a deficiency in porphyrin synthesis, which leads to sideroblastic anemia. On the peripheral smear, the red cells are hypochromic, but the presence of target cells and ring sideroblasts help differentiate this process from IDA. Characteristically, in toxic anemias the serum iron is elevated and the transferrin saturation is increased; the opposite is found in IDA. Examination of the bone marrow, if performed, reveals the presence of ring sideroblasts, along with increased iron stores common in porphyrin synthesis deficiencies.

Chronic diseases, such as malignancies, and inflammatory processes, including arthritis and the collagen-vascular diseases, may cause anemia, but usually the blood film reveals normochromic and normocytic cells. Unfortunately, the symptoms from these processes are protean, and they may be difficult to differentiate from those of IDA if the peripheral blood picture is similar. As with IDA, patients with chronic disorders will characteristically have low serum iron levels. Elevated-to-normal serum ferritins are seen in these disorders, while patients with IDA exhibit decreased ferritin values. In those affected with inflammatory processes, the transferrin saturation is decreased and UIBC is increased, as seen in IDA. On the other hand, patients with malignancies exhibit an increased transferrin saturation and decreased UIBC (see Fig. 3). Another screening device to differentiate IDA from acute and chronic inflammation is the erythrocyte sedimentation rate. It almost always is increased in neoplasia or acute and chronic inflammation over that found during normal pregnancy, whereas in severe IDA it is normal. Erythrocyte survival times are slightly decreased in IDA and may also be shortened in chronic diseases. Infestation (e.g., hookworms), although rare in this country, must be included in the differential diagnosis in developing countries, in world travelers, or in areas of poor sanitation. Other causes of excess intestinal iron loss, such as diverticulitis, intestinal cancer, and peptic ulcer disease, are rare during the menstrual years and even more infrequent during pregnancy.

Pyridoxine-responsive anemias are characterized by hypochromic, microcytic erythrocytes with increased ring sideroblasts and prominent target cells. Unlike IDA, serum iron levels, transferrin saturations, and serum ferritins are increased in patients with pyridoxine-responsive anemia. The UIBC is characteristically decreased.

In hemoglobin synthesis abnormalities, such as heterozygous thalassemia, a microcytic, hypochromic pattern may also be demonstrated, and these abnormalities may be confused with IDA. RBC morphology usually shows more basophilic stippling and target cells in thalassemia patients; these are not common in patients with IDA. In a person of African or Mediterranean origin, thalassemia should be ruled out by hemoglobin electrophoresis. Other aids in diagnosis include the erythrocyte count, the MCV, and the RC. Almost 85% of patients with heterozygous thalassemia have an RBC mass greater than 5 million/mm3 despite a reduced hemoglobin concentration. In contrast, only 3% of adults with IDA have RBC counts over 5 million/mm3 The MCV is reduced in heterozygous thalassemia; values of 55 to 70 cu μ per cell are the rule, whereas values below 70 cu μ per cell are very uncommon in IDA. Reticulocytosis of greater than 5% is much more commonly noted in thalassemia patients than in those with IDA. In addition, serum iron concentration and transferrin are not usually altered in the hereditary anemias; however, serum ferritin levels are characteristically normal to increased.

In the final analysis, the response to iron therapy may prove to be diagnostic. This method is commonly used in the low-risk patient and can be harmful only in patients who have allergic reactions or in those who have iron overload due to other hematologic diseases. In any case, in “therapeutic” diagnosis of IDA, the patient should be followed carefully to detect iron-unresponsive anemia. If the patient with IDA receives sufficient medication, an intense reticulocytosis should occur between the seventh and tenth day after the initiation of therapy. A significant increase in hemoglobin values should be evident in 3 to 4 weeks, and the hemoglobin concentration should approach normal values within 2 months. This may not occur during pregnancy, however, since the increase in RBC mass and transfer of iron to the fetus may continue the iron-depletion process. In this case, if other factors such as inflammatory diseases, chronic infections, or hemoglobinopathies are not suspected, the iron therapy should be continued until the end of the pregnancy, since eventually there will be a response.

Iron Therapy.

Iron can be administered as elemental salt (orally), as an iron-carbohydrate complex (parenterally), or as a blood transfusion. The oral route is preferable. Table 5 lists the common iron preparations used for iron supplementation and the cost of these compounds. The ferrous sulfate preparations are the least expensive and the most logical choice for most patients.11 Other iron salts or forms of ferrous sulfate are no better absorbed and are more expensive. Some iron preparations have been combined with vitamins. These complexes are likewise no more effective than simple iron preparations and are more expensive. This, plus the fact that they are usually given in a one-a-day dosage, makes this therapy as a singular form of iron supplementation an unlikely choice. One of the principal uses of these combination preparations, however, might be in the iron-sufficient woman who exhibits evidence of good iron stores, in whom daily administration of iron might be sufficient. Nevertheless, hematologic assessment of the patient's iron status during the second and third trimester needs to be completed to determine if this amount of iron is adequate.

TABLE 5. Iron Supplementation Preparations


Cost per 100

Retail Cost

Tablets or Capsules

for 9 Mo

Ferrous sulfate



Fer-In-Sol caps



Feosol tablets



Ferrous gluconate












Slow FE



Feosol elixir

$10.99/16 oz


Ferrous sulfate elixir

$6.29/16 oz


Parenteral Imferon




1 unit


NA, Not available.

Sustained-release capsules were previously thought to be poorly absorbed. More recent data using refined techniques have shown the reabsorption to be 80% normal; therefore, consideration of their use in patients who may be iron intolerant owing to gastrointestinal side-effects is reasonable. In addition, slow-release iron tablets may provide a reasonable alternative in these parturients. For other patients who experience intestinal distress, ferrous sulfate liquid can be used. This preparation should be well diluted in juice or water so that staining of the teeth does not occur.

Most of the oral agents contain 200 to 300 mg of iron complex, yielding approximately 40 to 60 mg of elemental iron.11 Likewise, the ferrous sulfate syrup (40 mg/ml) or the pediatric drops (120 mg/ml) yield 20% of this dose in elemental iron. The ideal dosage is 60 mg of elemental iron administered three to four times daily. The value of the spaced dosage is related to the absorption rate, which is maximal about 4 to 6 hours after each dose. Most commonly, the iron is administered with meals during pregnancy, since this seems to reduce the incidence of gastric irritation. Other side-effects, such as diarrhea and constipation, occur in approximately 5% and 15% of pregnant women, respectively. However, diarrhea has been reported in about 3% and constipation in as many as 10% to 12% when patients are given placebos. Therefore gastrointestinal problems may be related to pregnancy and not to iron therapy.

It is recommended by some that oral iron be used in pregnancy regardless of the hemoglobin status because iron loss due to pregnancy will certainly exceed the amount obtained in the diet.12 Support for iron supplementation in the nonanemic pregnant woman for the prevention of iron depletion comes from nutritional surveys demonstrating that diets in the United States usually contain an average of 5 to 10 mg of iron daily. Unfortunately, iron is better absorbed from meat and dairy products than from cereals and most vegetables. Since much of our population, including the impoverished, subsist on a high percentage of processed, gluten-positive food, it is doubtful that a woman receives adequate iron during pregnancy. A food fortification program for the prevention of IDA has been suggested in which iron would be added to bread, flour, and cereals. However, some of these sources may not be well absorbed. An increase in hemochromatosis in iron-sufficient persons (nonpregnant) receiving iron supplementation in their food has been reported in studies from Scandinavia; this finding has not been reported in populations in which IDA is common. Finally, when one views the cumulative effects of a diet chronically low in iron, adolescence, previously excessive menstrual loss, and early successive pregnancies, almost every pregnant woman in our society appears in need of supplemental iron in therapeutic doses. Certainly all pregnant women with PCV less than 35% to 37% and hemoglobin levels below 12 g/dL should receive iron, particularly if other laboratory tests support the diagnosis of iron deficiency or depletion. The therapy usually is continued throughout pregnancy and during the postpartum period. For the patient with IDA, progress should be monitored, and if the hemoglobin level is unchanged after 6 weeks of therapy, other hematologic problems should be considered.

Treatment with parenteral iron is rarely necessary unless the patient with IDA cannot or will not take the iron by the oral route, has an absorption disorder, or must undergo a surgical procedure within 1 to 2 weeks. Parenteral iron is available as a complex with either dextran or sorbitol and can be given intramuscularly or intravenously.13 Minor side-effects such as skin staining, discomfort at the site of injection, pruritus, malaise, and a metallic taste occur in 8% to 10% of patients. Moderate reactions such as lymphadenopathy, phlebitis, severe headaches, hyperpyrexia, and leukemoid reactions occur in 1% to 2%. Severe systemic adverse reactions such as anaphylaxis, renal toxicity, and bronchial spasm occur in approximately 0.5% of patients. The incidence of these reactions can be reduced by eliminating those patients with a history of eczema, asthma, or other allergic phenomena.

When parenteral agents are administered, a test dose of 0.5 ml should be given and the patient observed for 15 to 30 minutes before the remainder of the dose is given.13 Dosages of 5 mg/kg or 250 mg per each gram of hemoglobin less than 14 g/dL are standard. The rate of infusion for the intravenous administration should not exceed 50 mg/minute. There have been no adverse experiences associated with the human fetus or newborn. Infants of mothers who were treated with parenteral iron during the course of gestation, just prior to delivery, or during lactation showed no signs of hemochromatosis.

Severe anemia, particularly in the patient with acute hemorrhage, is probably best treated by transfusion. Transfusion obviously provides an immediate improvement in the anemia and is a ready source of iron for the production of new RBCs. Transfusion is not often used in the treatment of IDA owing to its expense and to the potential for serious adverse effects (i.e., hepatitis, human immunodeficiency virus (HIV), and transfusion reaction). It is useful in patients with IDA who are undergoing major surgery and who require immediate replacement of RBCs. In the normal postpartum patient, transfusions are rarely needed unless the PCV is less than 20% to 25% or the subject is symptomatic, since most patients can be adequately treated with oral iron therapy.

In summary, patients with iron depletion or deficiency will rarely be detected prior to changes in the hemoglobin or PCV level. Supplemental iron therapy is recommended in most pregnant women because (1) iron depletion is so common, (2) therapy is safe and inexpensive, (3) lack of response to supplements is an indication for further testing, (4) other testing is avoided, in most cases, and (5) the effects of an undetected common disorder like IDA on the mother and fetus are not fully known. Therefore, unless more severe hematologic disease is suspected, a therapeutic trial of iron should be given, at least to those at risk for iron depletion, if not to all pregnant women.14 Finally, diagnosis of almost any anemia may be handled on an ambulatory basis. However, there should be active hematologic consultation when the diagnosis is unsure, severe disease is suspected, or the perinatologist is unfamiliar with the diagnostic requirements.


Another disorder in hemoglobin production is thalassemia, which is caused by an alteration in the rate of synthesis of the α- or β-chain. The hemoglobinopathy is classified based on which chain is affected. Clinical symptoms may also be used to classify thalassemia as thalassemia major, intermedia, and minims. In general, only the homozygous forms are in the major category, while the heterozygotes (α, β, α-β , δ )demonstrate variable degrees of symptomatology.15

Homozygous Forms.

α-Thalassemia appears to be most severe, with fetal death occurring in most cases. The poor outcome is due to the increased production of α-chains, which combine to form Bart's hemoglobin, a high-oxygen-affinity hemoglobin that functions poorly for oxygen transport. Electrophoresis on fetal blood shows 85% to 100% Bart's hemoglobin, with only small amounts of hemoglobin A and hemoglobin F. The infant develops severe anemia and hydrops similar to that noted in those with Rh isoimmunization. Homozygous α-thalassemia is the most common form of hydrops fetalis in Southeast Asia, but is rare in this country. Homozygous β-thalassemia (Cooley's anemia) is characterized by severe anemia, usually leading to death in early childhood. On electrophoresis, the hemoglobin is usually 40% to 70%; hemoglobin A2 is mildly increased, while hemoglobin A is reduced and varies with the severity of the anemia. Treatment involves repeated transfusions, and pregnancy is infrequent since the patient usually dies in childhood.

Heterozygous Forms.

Patients heterozygous for thalassemia rarely manifest signs or symptoms of the disease. Heterozygous β-thalassemia is the most common form; patients usually exhibit only mild hypochromic, microcytic anemia, with elevation of the hemoglobin A2 (above 3.5%) and a mildly increased hemoglobin F percentage (2% to 5%). Heterozygous α-thalassemia is difficult to detect by laboratory means because the affected globin chain is common to all types of hemoglobin. Therefore, there is a reduction in the amount of hemoglobin A, F, and A2, while the percentage of these compounds remains the same as in normal persons. This finding, combined with the fact that anemia is usually mild, also complicates the detection process. Although family pedigrees or intricate hematologic studies are usually required for a firm diagnosis in adults, α-thalassemia is easier to delineate in the neonate, since Bart's hemoglobin and small amounts of hemoglobin H (four β-chains) may be present. Concomitant pregnancy reveals few adverse maternal or fetal effects, except a slight increase in spontaneous abortions. In general, the maternal outcome is related to the severity of the anemia; those more severely affected have an increased maternal morbidity, although fertility does not appear to be affected. The double heterozygote (α-β-thalassemia) is also difficult to diagnose and may be more common than previously suspected. Pregnancy statistics are sparse, although most authors suggest that the anemia and other symptoms are more severe than with α-thalassemia or β-thalassemia trait alone.

Prenatal Diagnosis.

The prenatal diagnosis of thalassemia syndromes should be offered to any couple with a prior affected child or those at risk because of ethnic origin or pedigree. Prospective parents not having an affected child previously need verification of carrier status based on MCV, zinc protoporphyrin, hemoglobin electrophoresis, and pedigree analysis before fetal assessment is undertaken. A sample of fetal deoxyribonucleic acid (DNA) can be obtained by either chorionic villus sampling, amniocentesis, or percutaneous umbilical (cord) blood sampling. DNA analysis of that portion of the α-globin or β-globin gene containing the deletion or mutation can be identified using restriction fragment length polymorphisms and Southern blotting techniques. More recently, direct DNA sequence analysis has become available by performing polymerase chain reaction amplification of fetal DNA. Using this method of detection, prenatal diagnosis can usually be confirmed within 7 days of fetal sampling.16


Management of the thalassemia traits during pregnancy is much the same as for the patient with normal hemoglobin. Proper diagnosis is difficult, since the mild microcytic, hypochromic anemia can be easily confused with IDA, a condition that frequently coexists. Folic acid supplementation is recommended because of the increased utilization and high erythrocyte turnover. Blood transfusion or other invasive therapy should be avoided, if possible, unless severe anemia or other problems intercede.

Disturbance in Maturation: DNA Synthesis (Megaloblastic Anemia)

Although not specific, the term megaloblastic anemia is used by most hematologists to describe a group of hypoproliferative disorders that have a characteristic morphologic appearance, ineffective erythropoiesis, and a moderate hemolysis of the circulating RBCs. Pernicious anemia (lack of vitamin B12) and folate deficiencies are the prototypes of this disorder. The underlying biochemical defect is impaired thymidylate formation, an essential rate-limiting initial step in the DNA synthesis of the body's cells, which require tetrahydrofolic acid as a coenzyme.

The megaloblastic changes occur as a result of a maturation defect in the marrow affecting erythrocytic, leukocytic, and thrombocytic cell lines. The lack of vitamin B12 or folate slows DNA synthesis and delays the nuclear maturation of chromatin of the immature cell into the dense cyanotic figure associated with the normoblast during normal erythropoiesis (see Fig. 2). Although the nucleus remains large and immature, the cytoplasmic mass decreases normally as it does during maturation. Therefore, the large macrocytes found in the peripheral blood of those with megaloblastic anemia represent a delay in nuclear maturation in the erythrocytic series and are diagnostic by their abnormal nuclear cytoplasmic appearance. Likewise, in the leukocytic series, the abnormal granulocyte is referred to as a giant metamyelocyte because of the condensation failure in the large nucleus. Due to granulocyte marrow turnover, leukopenia develops and the cells found in the peripheral blood are hypersegmented, with the nucleus having six or more lobes rather than the normal three to five. Finally, inadequate thrombopoiesis occurs, which results in an increase in bone marrow megakaryocyte mass with fewer platelets in the peripheral blood. The platelets that are present may function poorly, and a bleeding diathesis may be present if the vitamin B12 or folate deficiency is severe.


FAD appears to be the most common cause of megaloblastic anemia. However, vitamin B12 appears to catalyze the conversion of 5-methyltetrahydrofolate to its active form, as well as to affect the storage and transport of folic acid within the body. Therefore, a deficiency in vitamin B12 may lead to megaloblastic anemia singularly or by its effect on folate metabolism. Causes of folate or vitamin B12 deficiency include decreased intake and conditions in which requirements are increased (e.g., pregnancy, hyperthyroidism, hyper parathyroidism, and malignancies). Causes that are not related to vitamin B12 or folate therapy include the purine and pyrimidine synthesis inhibitors (cancer chemotherapy), pyridoxine responsive megaloblastic anemia, and Di Guglielmo's syndrome (erythremic myelosis), in which platelet production is snore severely affected than erythrocyte or leukocyte maturation. Finally, in various inborn errors of metabolism, such as Lesch-Nyhan syndrome and hereditary orotic acid deficiency, either a lack of enzymatic catalases or metabolic blockage to folate incorporation can cause megaloblastic anemia. All the infrequent causes are difficult to diagnose, and megaloblastic anemia due to these causes is unresponsive to folate therapy. Fortunately, vitamin B12 deficiency and FAD are the most common etiologies of megaloblastic anemia, accounting for 98% of reported cases. FAD is a much more common cause during pregnancy (99%), and it causes 92% of all cases in any age group.17 Vitamin B12 deficiency is extremely uncommon during pregnancy (1:8000 pregnancies); the lack of intrinsic factor almost always occurs in persons beyond the childbearing age unless significant gastric surgery has been performed.


The major clinical signs and symptoms of megaloblastic anemia are variable and are usually not detectable until the anemia is severe. As in other anemia processes, the symptoms increase in severity with the progression of anemia, from pallor, through weakness, malaise, dizziness, and shortness of breath, and finally to congestive heart failure. Patients with megaloblastic anemias may appear jaundiced more frequently than those with IDA because of the rapid cell turnover and their increased propensity for bleeding diathesis. If anemia is present, the appearance of the peripheral smear is usually diagnostic, demonstrating several hypersegmented granulocytes as well as RBC inclusions (i.e., stippling, Howell-Jolly bodies, Cabot's rings, and nonhemoglobin iron). The most sensitive indicator of early folate deficiency appears to be the hypersegmentation of neutrophilic nuclear material (average low value >3.27 or more than 4% with more than five lobes in 100 consecutive polymorphonuclear neutrophils). Serum folate values in pregnant women are usually lower than in the nonpregnant patient, and these decrease progressively toward term. In the absence of IDA, however, a low fasting serum folate (<3 ng/ml by radioimmunoassay) is virtually diagnostic of folate deficiency. In addition, a low erythrocyte folate activity (<20 ng/ml) is probably the best biochemical index of FAD. Although the anemia is macrocytic, it is usually normochromic with normal MCH and MCHC. The MCV, however, is strikingly elevated in contrast to IDA and may exceed 150 cu μ (see Table 1). The high MCV is helpful because it is in contrast to that found during pregnancy, IDA, and other microcytic processes. In general, the more severe the anemia, the more bizarre the erythrocytic changes, with nucleated RBCs appearing in the peripheral blood smear when the PCV is less than 20%. Unfortunately, if folate is deficient, there is usually a concomitant IDA, which may confuse the diagnosis.18 However, the RC is lower in a megaloblastic anemia compared with the level in IDA.

In contrast to findings in IDA, most megaloblastic anemias demonstrate an elevated serum iron saturation and reduced UIBC. These ferrokinetic changes are usually combined with elevated plasma bilirubin and uric acid levels, which demonstrates that the ineffective erythropoiesis is associated with intramedullary hemolysis, a classic finding in megaloblastic anemia. This hypothesis is supported by increasing levels of lactic dehydrogenase (LDH1 and LDH2), serum muramidase, and malic dehydrogenase as the severity of megaloblastic anemia progresses. Similarly, RBC survival time is only 50 to 60 days versus the normal 120, denoting a concomitant extramedullary hemolysis. In contrast, the liver is not involved; serum glutamic oxaloacetic transaminase (SGOT) and alkaline phosphatase are usually normal.

Folic Acid Deficiency.

Folic acid (pteroylmonoglutamic acid), like the essential amino acids, cannot be synthesized by humans and thus must be provided in the diet. Deficiency of this substance continues to be an infrequent cause of severe anemia, but more commonly represents a simple deficiency.17 This variable presentation probably reflects the ubiquitous presence of folic acid in meats and fruits, as well as in dried or fresh vegetables. Although not a common cause of frank anemia during pregnancy, folate deficiency has increased in incidence from 1:250 pregnancies in 1960 to 1:50 in 1975. This increase probably reflects better diagnostic methods and an increased index of suspicion, since no greater percentage of women are clinically anemic. However, it demonstrates very graphically that many pregnant subjects are potential candidates for folate deficiency.

Folic Acid Metabolism.

Pteroylmonoglutamic acid, or folic acid, is the parent compound for a large family of chemicals collectively known as folates. According to nomenclature developed in 1965, the molecule contains a pteridine derivative, a p-aminobenzoic acid (PABA) residue, and l-glutamic acid (Fig. 8). To be useful in the synthesis of purines and pyrimidines, folic acid must be converted to 7,8-dihydrofolic acid (FH2) and finally to the active form, tetrahydrofolic acid (FH4). The metabolically important members of the folate family are FH4 derivatives and differ by the addition of one carbon unit to the terminal nitrogen of the PABA unit. Three oxidation levels of this carbon are possible: formyl, hydroxymethyl, and formimino (see Fig. 8). The formation of formyl FH4 is catalyzed by the enzyme dihydroglycolate reductase. It is this portion of the metabolic pathway that the folate analogs (chemotherapeutic agents) inhibit. Formyl FH4 derivatives are important in DNA synthesis and the production of amino acids and the enzymatic catalyst citrovorum factor. Without the formyl FH4 derivatives, thymidylate cannot proceed in DNA synthesis, which ultimately causes the hematologic manifestations noted in megaloblastic anemia. Interference with the hydroxymethyl series, in contrast, impairs purine synthesis, although no clinical manifestations have been related to this disorder. Finally, inhibition of formimino group synthesis disturbs histidine catabolism, and while this has no adverse effects, it does provide the basis for tests useful in the diagnosis of FAD; formiminoglutamic acid is excreted in increased amounts, particularly when oral histidine is administered to folate deficient persons.

Fig. 8. Structure of folic acid, including its three essential derivatives.

Folic Acid Requirements.

The minimum daily requirement of folic acid or its equivalent is 50 μg/day in the normal adult or 100 μg/day in those with additional requirements (e.g., pregnancy or adolescence). Although the average diet contains many times this amount, most of it is unavailable. In addition, the body is thought to contain only approximately 5 mg of folate compounds total, so the reserves are much smaller than those of vitamin B12 or iron. For these reasons, the official recommended daily allowance of food folate for the adults is 0.4 mg/day; during pregnancy, 0.5 to 1 mg/day is considered adequate.

The richest vegetable sources of folate are asparagus, broccoli, spinach, lettuce, and lima beans, each of which contains more than 1 mg of folate per 100 g dried weight. The best fruit sources are lemons, bananas, and melons. The folate in most vegetables and fruits is conjugated, however, and must be broken down in its absorptive pathway through the gastrointestinal tract. Although high concentrations are found in the liver, kidney, and brain of carnivorous animals, these are poor sources of usable folate.

Several problems may prevent attainment of a positive folate balance. They include (1) failure to consume adequate amounts of food high in folate, (2) failure to prepare foods properly (most vegetables and meat are overcooked, resulting in folate destruction), and (3) increase in requirements (e.g., in adolescence, successive pregnancies, and multiple gestation). Women do not appear to have an increased need or an increase in daily loss compared with men. During pregnancy, however, approximately 100 μg folate is needed daily to supply the fetus and maintain maternal stores (twice normal nonpregnant requirements).17 The higher frequency of FAD during pregnancy and the puerperium is attributable to several factors. Most important is the elevated demand of this substance by the placenta, the fetus, and the expansion of the maternal RBC mass. In addition, compromised socioeconomic conditions, which are usually related to inadequate nutrition, may also be major factors. Finally, complications specific to pregnancy, such as unusual fad diets, prior gastrointestinal tract surgery or malabsorption states, prolonged hyperemesis, or the slowed gastrointestinal absorption of folate (thought to be due to increasing levels of estrogen and progesterone), all may give rise to an increased incidence of FAD. Chronic infection of any type, as well as other causes of anemia, the presence of multiple fetuses, or closely spaced pregnancies, as well as lactation, may also substantially increase maternal requirements for folic acid. Also, many drugs, such as ethanol and certain antibiotics (nitrofurantoin, trimethoprim), as well as the anticonvulsant diphenylhydantoin, may impair absorption or utilization of folate. Although it is an infrequent cause of anemia, folate deficiency should be detected and treated, since megaloblastic anemia will eventually result.

Folic Acid Absorption.

Although the mechanism of folate absorption in the intestine is unclear, the proximal jejunum seems to be the principal site of uptake. When radioactive labeled folic acid is administered in the unconjugated form, only the monoglutamate form (FH1) appears in the plasma; thus, folic acid or conjugated derivatives in food products must be broken down before being absorbed. The intestine contains several pancreatic conjugates to reduce the FH4 to FH1, the site of degradation being within the lumen of the intestine or near the brush border of the intestinal cell. After entry into the serum intact, folates, as well as their cleavage products, are excreted by the liver into the bile. Although some of this folate may be reabsorbed, the loss of folate by bile secretion may accelerate depletion in patients with mild absorptive problems.

As shown in Fig. 9, once FH1 enters the mucosal cells, it can be absorbed against a concentration gradient, although a passive transport may also occur. The amount of FH1 absorbed in the mucosal cells depends on the need, which is reflected by the content of FH4 within these cells. If the level is high, indicating adequate folate stores, little of the hydrolyzed FH1 will be transported, whereas under conditions of decreased FH4 availability, the FH1 is quickly absorbed into the mucosal cells. Up to 90% of the FH1, even in deficiency states, is reduced to FH4 and methylated in the mucosal cells before it is released into the circulation. Approximately 60% is cleared in one circulation, and 90% to 95% is cleared within 3 minutes. The rapidity of clearance suggests that folate is bound quickly by a serum protein that is increased in FAD. The elevation in this binding protein also occurs in patients with uremia and in women who are pregnant or are taking oral contraceptives. It appears that this moiety consists of two proteins, with molecular weights of 200,000 and 50,000, respectively. The larger fraction is present in human milk and leukocyte membrane. It appears to have antibacterial properties similar to those of the β-lactoglobulin. The smaller portion, however, behaves as a β-globulin similar to transferrin and participates in the binding and release phenomena. After release to a metabolically active cell, the folate is held in storage or used for biosynthesis.

Fig. 9. Absorption of folic acid from the intestinal lumen through the mucosal epithelium and into the mesenteric circulation.

Clinical Features.

Clinical syndromes of FAD include all of the nonspecific features described earlier in the section, Anemia Detection. The most notable features include roughness of the skin and glossitis, which are seen with both vitamin B12 and folate deficiency (Plummer-Vinson syndrome). It is important to be certain that no neurologic deficiency is present, since the neurologic abnormalities of vitamin B12 deficiency are not remedied by folate supplements. Specific laboratory diagnosis of folate deficiency includes the use of peripheral blood smear, PCV, hemoglobin values, and RBC indices. Serum folate assay, a microbiologic procedure employing the organism Lactobacillus casei, is a reliable method for diagnosis of FAD. Although serum folate is slightly lower in pregnant subjects (see Table 1) compared with normal subjects, it does not approximate folate levels found in megaloblastic anemia patients, which are commonly lower than 3 ng/ml. However, since the serum folate level falls many weeks before anemia appears, a level lower than 5 ng/ml signifies a deficiency of folate that leads to anemia if not treated.18 Although white blood cell and RBC folate assays can be performed and may give a better assessment of tissue folate, the assays are difficult to perform, hard to standardize, and not widely used. Measurement of formiminoglutamic acid excretion in response to histidine loads is less specific than serum folate measurement, but it gives a better measure of folic coenzyme levels than does the microbiologic assay. Unfortunately, it is difficult to quantitate because histidine absorption and excretion are altered during pregnancy. Finally, the response to therapy is useful clinically. As shown in Table 6 and Table 7, changes in the above parameters occur quite rapidly if anemia is due to folate deficiency. Since the deficiency of vitamin B12 is rare in pregnancy, folate therapy in the absence of lowered vitamin B12 levels appears to be justified in most cases.

TABLE 6. Changes in Laboratory Values Related to Duration Folate Deficiency

Laboratory Assessment

Duration (wk)

Low serum folate


Abnormal neutrophil


Increased formiminoglutamic acid


Low folate (RBC)






TABLE 7. Response to Folate Therapy in Folic Acid Deficiency


Response Time

Reticulocytosis, thrombocytosis

48 h

LDH, LDH2 decrease

96 h

PCV and hemoglobin increase

1 wk

Neutrophil normal

2 wk

Normal folate

6 wk

Differential Diagnosis.

Conditions that may result in malabsorptive syndrome include subtotal gastrectomy, nontropical sprue (gluten-induced celiac disease), tropical sprue, lymphomatous infiltration of the small intestine, Whipple's disease, scleroderma, amyloidosis, and diabetes mellitus. Various drugs, such as oral contraceptives and anticonvulsant medications (hydantoin derivatives), may also interfere with folate absorption. Conditions that increase folate requirements and thus can result in folate deficiency include pregnancy, hyperactive hematopoiesis (hemoglobinopathies and hemolytic anemias), malignancies, and various skin diseases such as psoriasis and chronic exfoliative dermatitis. Finally, the activation of FH4 may be blocked in patients treated with folic acid antagonists and in patients with scurvy.


Folic acid is usually administered orally on a prophylactic or therapeutic basis as a 1-mg tablet, although this dose is greater than the requirement. Moreover, oral folate is given once a day, and since folate absorption, even in deficiency states, is only 10 μg to 20 μg every 6 to 8 hours, the amount of folate absorbed may be minimal. More recently, the 1-mg tablet has been given as one half tablet (0.5 mg) two to four times a day for prophylaxis or treatment of a true folate deficiency. Should neurologic signs be present, serum B12 levels should be measured prior to folic acid therapy, because folate treatment corrects the anemia but not the neurologic defect. The principal indication for parenteral folic acid occurs when specific folic acid antagonists must be used. Toxicity from folic acid has not been observed, even in doses hundreds of times the usual therapeutic level, and only rarely have sensitivity reactions been recorded. Indications for folic acid therapy are (1) actual or anticipated deficiency (e.g., in multiple gestations and adolescent pregnancy), (2) prophylaxis for women with poor nutrition and other high risk factors, and (3) women taking certain drugs (e.g., trimethoprim, diphenylhydantoin) on a long-term basis.

In the patient with actual megaloblastic anemia, one can expect an elevation in the RC approximately 48 to 72 hours following therapy (see Table 7). This appears to be the earliest sign of remission and is usually dramatic. The thrombocytopenia that may be present rapidly clears within the first few days. The specific lactic dehydrogenase (LDH) isoenzymes also decrease early in the course of therapy. A slight increase in PCV and hemoglobin level may be evident by the end of the first week, and the neutrophils gain their normal appearance after approximately 2 weeks. Serum iron levels may have been falsely elevated in folate deficiency because iron stores were not efficiently used for erythropoiesis; these usually fall to subnormal levels, indicating a coexisting IDA during the process of folate replacement.

Folate Deficiency During Pregnancy.

Maternal folate deficiency is common in late pregnancy. In England and Australia, 95% and 60%, respectively, of women at term have been reported to have low serum folate levels.19 Sixty percent of women in the United States also demonstrated low serum folate levels. Although frank megaloblastic anemia of pregnancy is less frequent, studies in Canada have shown approximately 25% of pregnant women have folate deficiencies sufficient to produce some megaloblastic changes.17 Anemia cannot be completely overlooked, since one study reported that almost 500 women in 3200 obstetric patients had a hemoglobin level below 9.5 g/dL and 90 of these had frank megaloblastic anemia.17 Embryopathies, especially neural tube defects, have been associated with folate deficiency and supplementation appears to decrease the percentage of such complications.20 FAD has been associated with low-birth-weight infants, smaller maternal blood volume, abruptio placentae, prematurity, and other maternal-fetal disease processes. However, other studies, particularly those with specific assay techniques for folate deficiency, have not shown a causal relationship.19 Indeed, this association tends to occur in women who are, in general, nutritionally deprived. Since the effects of deprivation are unclear, most physicians favor supplementation for patients at high risk for folate deficiency.

Vitamin B12 Deficiency.

Vitamin B12 is a unique compound, since it is synthesized only by certain microorgansims. High concentrations of microorganisms producing vitamin B12 are found in liver, seafood, meat, eggs, and milk. Vitamin B12 is enzymatically important as a cofactor in the formation of succinyl-CoA and in the folate pathway for the production of active FH4. It is in these two enzymatic pathways that vitamin B12 deficiency is clinically apparent. The general clinical features of vitamin B12 anemia are the same as for any other anemic process, except for the peripheral neuropathy characteristic of this disorder. Thus, the physical and laboratory examinations are more informative than the history.


Intrinsic factor is necessary to facilitate absorption of vitamin B12 in the ileum. Intrinsic factor has been shown to be an alkaline stable glycoprotein secreted by the parietal cells from the fundic mucosa of the stomach. Secretion of intrinsic factor parallels that of hydrochloric acid and therefore is concomitantly reduced in achlorhydric patients. Vitamin B12 derivatives are bound to intrinsic factor in a complex that attaches to specific mucosal receptors on the microvilli of the ileum. Other secretions such as pancreatic enzymes may promote vitamin B12 absorption by protecting the receptor sites and providing a suitable environment for intrinsic factor—vitamin B12 complex. The complex partially enters the cell by pinocytosis and dissociates near the membrane, leaving the vitamin B12 in the cytoplasm. Uptake from the mucosal cell into the serum is slow and requires active transport.


Various laboratory tests, such as measurement of the vitamin B12 level, can be performed to distinguish this deficiency from megaloblastic anemias due to other causes. It should be remembered that the serum folate level may be artificially lowered in vitamin B12 deficiency even though folate stores are adequate; thus, serum folate may be decreased while RBC folate is normal. Laboratory methods for measurement of vitamin B12 use fecal and urine samples (e.g., the Schilling test), which involves giving a small dose of radioactive cobalt; thus, it is not used during pregnancy. The measurement of serum vitamin B12 may be performed by radioactive assay, and although maternal serum levels fall progressively during gestation to intermediate levels (80 to 120 pg/ml), the finding of levels below 50 pg/ml is very suggestive of pernicious anemia.


Appropriate therapy includes six weekly injections of 1000 μg hydroxycobalamin. Further diagnostic tests with augmented histamine or the Schilling test are usually performed during the postpartum period. In addition, standard prophylactic iron and folic acid supplementation should be implemented. It is particularly necessary to treat postpartum women who are breast feeding. Indeed, the first signs of severe vitamin deficiency may occur in these women.

Effects on Pregnancy.

Pernicious anemia usually does not complicate pregnancy. The incidence of vitamin B12 deficiency during pregnancy is 1:6000 to 1:8000. It is usually associated with a strict vegetarian or other fad diet, incipient pernicious anemia due to lack of intrinsic factor, or a chronic gastrointestinal disorder such as tropical sprue. Women who have pernicious anemia are usually infertile and are usually not of reproductive age. In healthy women, vitamin B12 deficiency takes 2 to 3 years to develop, even with a severely restricted diet. Since a pregnancy only lasts 9 months, with a peak requirement for vitamin B12 of less than 4 months, it is difficult to develop pernicious anemia during a pregnancy. During pregnancy there is a steady fall in the serum vitamin B12 level due to the rapid expansion of PV and the increased utilization of vitamin B12 by the fetus during the latter half of pregnancy, accounting for the reported 15% incidence of low serum levels of vitamin B12 during pregnancy. Under other circumstances, these women might be considered to have vitamin B12 deficiency. Nevertheless, if the diagnosis is suspected but cannot be confirmed, vitamin B12 therapy should still be administered. If the serum level does not return to normal within 6 weeks after therapy, it is advisable to perform other investigations, particularly if other adverse symptoms are present.

Bone Marrow Failure

The failure of bone marrow to produce sufficient erythrocytes is a rare occurrence in obstetric patients. Nevertheless, the resulting pancytopenia, which affects all cell lines, is devastating. A selective aplastic erythropoietic cell line is called pure red cell aplasia. In children, the inherited type of this disorder is called Diamond-Blackfan syndrome; a pregnant woman, who had Diamond-Blackfan syndrome as a child, transmits the trait to 25% of her offspring, who are born with minor congenital malformations. Acquired red cell dysplasia is associated with thymic tumors in 30% to 50% of cases and occasionally occurs during pregnancy. This disorder usually is thought to have an immunologic basis and presents as a normochromic, normocytic anemia with absolute reticulocytopenia as well as a normal leukocyte and platelet count. Bone marrow in this examination is normal except for erythroid hypoplasia. Treatment is usually by transfusion, although removal of an enlarged thymus is sometimes necessary. Although there have been few reported cases during pregnancy, the outcome is usually good.

Other disorders associated with bone marrow failure usually involve the replacement of bone marrow with either fatty, degenerative changes (aplastic anemia) or granulomatous/tumor infiltration (myelophthisic anemia).21 The development of aplastic anemia when there is fatty, hypocellular bone marrow is usually caused by exposure to drugs, chemicals, radiation, or several various diseases. Those agents known to be incriminated include acetylsalicyclic acid, chloramphenicol, gold salts, phenylbutazone, and tolbutamide. Both hepatitis virus A and B as well as total body irradiation have also been incriminated. The diagnosis is made by the appearance of pancytopenia on the peripheral blood count; normochromic, normocytic anemia; and reticulocytopenia. Diagnosis is confirmed by hypocellular or aplastic marrow appearance and fatty replacement of the bone marrow. The appearance in the blood cell smear on peripheral smear of nucleated red cells is more compatible with dysfunction rather than hypoplasia. A total granulocyte count of less than 200/ mm3 correlates closely with poor prognosis, since these patients are very susceptible to infection. In general, the basic goal of treatment is removal of the causative agent, if present, and supportive therapy to obtain remission if no cause can be found. Transfusion with packed red cells, platelet buttons, and granulocytes to maintain these elements at near-normal levels is recommended during the gestation. A variable effect has been noted with steroid administration; therefore, a therapeutic trial of steroids for 2 to 4 weeks is usually advised. In the nonpregnant state, stimulation with certain androgens such as oxymetholone (3 to 5 mg/kg/day) has been helpful. Bone marrow transplantation may be a therapeutic option in severe cases of this disease, but it depends on human leukocyte antigen (HLA) compatibility of sibling donors. Aggressive treatment with antibiotics and early identification of infection are extremely important, owing to the propensity of these patients to become infected. For that reason, antibiotic prophylaxis for delivery or operative procedures is recommended. Exposure to infected persons or a reduction in an environmental exposure to infections is recommended. In the hospital, reverse isolation can be used but is generally recommended only when the absolute granulocyte count is below 200 mm3.

In general, the outcome of pregnancy is good, particularly if the aplastic anemia has been diagnosed before pregnancy and is in a stable state. If the hemoglobin count is above 8 g/dL, the outlook is good. In addition, if the granulocyte count is above 600/mm3, the likelihood of infection is reduced. The major causes for maternal death remain hemorrhage and infection in 90% of patients. In general, the survival rate of the neonate is about 75%; prematurity is the major form of morbidity. Antiplatelet antibodies can be transported across the placenta, and in a large series, six infants were reported to be born with pancytopenia and a platelet antibody. For this reason, fetal platelet counts obtained by scalp sampling early in labor may be helpful.

Bone marrow infiltration by tumor or granulomatous cells (myelophthisic anemia) results in severe normochromic, normocytic anemia. The peripheral smear reveals a large number of fragmented cells with basophilic stippling reminiscent of a hemolytic process, although the RC is reduced. The white count, on the other hand, is elevated, and usually a left shift is noted in the differential. The combination of immature myeloid cells and normoblasts in the peripheral smear is a hallmark of this disease. Pregnancy among those with this disorder is rare, and reported cases have included metastatic disease from ovarian as well as bowel and hepatic cancer. The fetal and maternal outcome, of course, is dependent on the primary disease process.

Unknown Causes

The anemias associated with inflammation are usually chronic and are not common, since these are unlikely in the reproductive age group. Most of the disorders in this group occur from chronic suppurative infections or neoplasia. Patients with these disorders have adequate stores of iron (present as hemosiderin or ferritin), but the release mechanism from transferrin to the red cells and the iron pickup from the storage depot are abnormal. Characteristically, the anemia is mild: a hemoglobin of 8 to 11 g/dL and a PCV of 25% to 33% are the rule.22 In about 75% of cases, the anemia is normochromic, normocytic, but hypochromia and microcytosis can be manifest if the iron release is severely disturbed. As a point of differentiation, the UIBC is increased with a decrease in transferrin saturation with inflammatory processes. The opposite findings are seen in those patients affected with neoplasia. Serum iron levels are reduced in both of these entities. The RC is usually below 1%, while the sedimentation rate is elevated, frequently above that seen in pregnant patients. Diagnosis of the specific disorder is usually based on careful history and physical examination; a high index of suspicion is needed. Many of these patients, if pregnant, present to the perinatologists with iron-unresponsive anemia and may have subtle symptomatology such as a low grade fever or other signs out of proportion to the mild anemic process. Obviously, the diagnosis is assured by examination of the storage iron compartment, which will reveal sufficient quantities of iron; bone marrow assessment may also be useful in these cases. Therapy should be directed toward the specific disease process.

Back to Top

Anemias due to increased loss of erythrocytes can be divided into acute and chronic types. The chronic types exist in patients with intestinal disease, such as parasite infestation, significant hemorrhoids, or peptic ulcer disease. Since the loss of blood is chronic, the actual number of RBCs lost is probably not as important as the amount of iron lost in excess of what is absorbed. This, coupled with the increasing demand of the fetus and expansion of the maternal blood volume, can convert mild iron depletion to IDA during pregnancy. Although these causes are relatively rare in the reproductive age group in this country, careful assessment of iron unresponsive anemias in women from developing countries or with histories suggestive of intestinal disorders should be undertaken. Diagnosis and treatment are directed toward the specific disorder.

Anemias resulting from acute blood loss during pregnancy usually have evident etiologies, since external blood loss usually occurs and symptoms are sudden. These disorders can include multiple trauma and spontaneous splenic rupture, as well as disorders of the gastrointestinal, pulmonary, or urinary tract, which may or may not be related to obstetric conditions. Obstetric disorders may occur during early pregnancy (e.g., abortion, ectopic pregnancy, molar pregnancy) or in late pregnancy (abruptio placentae and placenta previa).

Blood loss at delivery may also account for significant losses of iron. Estimation of these losses has been measured with dye and isotope techniques, but large errors are common. Factors that lead to an underestimation of external blood loss at delivery include failure to take placental blood volume into account, incomplete measurement of hemoglobin in solution, incomplete extraction of hemoglobin from clots, and hidden bleeding. The average loss, determined by Crtechniques, is between 150 and 250 ml, depending on the patient's parity and the use of episiotomy, while an additional 30 ml is usually sequestered in the placenta. Therefore, approximately 200 ml of packed RBCs, equivalent to 500 ml of whole blood, is lost at the time of delivery. Approximately 230 mg of the 500 mg of iron used to expand blood volume is lost, while the remaining 270 mg is returned to the iron stores.

The placental “loss” of iron is variable, averaging 25 mg per pregnancy. The transfer of iron appears to take place by the same iron transferrin complex used in the mother. The placenta, however, appears to function independently, since it stores iron as ferritin and hemosiderin even in the absence of a fetus and after fetal death. The transport of iron across the placenta is unidirectional and preferential, with the fetus accepting iron even when maternal levels are low. This maternal fetal transfer involves an active transport mechanism, since the saturation of transferrin and the plasma iron level are considerably higher in the fetus. Most of the iron transferred across the placenta (70% to 80%) goes to the fetal liver, which serves in a storage capacity as well as a hematopoietic organ, The placenta also functions as a barrier to iron overload; thus, iron toxicity in the fetus by maternal overdose does not occur. Study of the incorporation of placental iron into fetal hemoglobin using radioactive material cannot be undertaken because in early investigations, use of labeled iron was associated with infant malignancies. However, indirect evidence shows heme incorporation into fetal RBCs to be similar to maternal utilization. Diseases in the mother (e.g., infection or inflammation) may decrease the amount of iron transferred, owing to poor storage mobilization and transferrin release. On the other hand, anemia due to iron loss during lactation, FAD, or chronic blood loss does not affect transfer of iron across the placenta.

The iron loss in lactation is small and usually averages 1 mg/day. Breast milk contains an iron binding protein similar to transferrin, called lactoferrin. This protein is synthesized in the mammary gland and stored in the ductal lining cells. The secretion of lactoferrin into the breast milk accounts for the iron loss noted in iron sufficient women during lactation. This compound is also bacteriostatic, and thus assists the infant in combating early infection and possibly prevents maternal breast infections. Moreover, since many lactating women do not have menses, the excretion of 1 mg of iron per day is offset by the avoidance of menstrual loss. Therefore, dietary iron supplementation as prescribed during and after pregnancy is usually sufficient to replenish iron stores during lactation.

During the postpartum period, acute bleeding problems may also occur owing to uterine atony, genital lacerations, or retained placenta. The acute causes, both obstetric and nonobstetric, are usually best treated with blood transfusion and removal of the offending agent. Obviously, the diagnosis depends on the specific disorder, as does the proper therapy. After the correct diagnosis is made and acute therapy is started, treatment with oral iron is usually indicated. Patients with acute blood loss usually show a normochromic, normocytic cell pattern with normal blood indices, hemoglobin levels, and PCV until after equilibration occurs, a process that may take from 12 to 24 hours. The iron studies are normal. Chronic blood loss, on the other hand, may be subtle enough to reveal only a loss in iron content if the loss in blood cells did not exceed the capacity of the bone marrow to produce the RBCs. In these cases, the chronic disorders would involve the same laboratory findings as IDA.

Back to Top

Diseases caused by increased destruction of erythrocytes present a hematologic profile characterized by persistent reticulocytosis, a finding that would indicate that there is increased erythrocyte destruction in excess of compensatory erythropoiesis by the bone marrow.23 The etiology of these disorders can be (1) extrinsic hemolytic anemia in which a mechanical, infectious, or immune factor is responsible for the decreased life of the red cell; (2) intrinsic hemolytic anemia in which the red cell loss is due to an inherited defect in the blood cell or membrane; or (3) unknown.

Extrinsic Hemolytic Anemia

Extrinsic hemolytic anemia is a proliferative disorder that involves erythrocytes that are structurally and metabolically normal but that, owing to environmental factors, suffer early death. Most frequent causes are hypersplenism, Coombs-positive hemolytic anemia, and microangiopathic anemia.


Enlargement of the spleen usually leads to excessive RBC sequestration and destruction and is most common in its severe and fatal form during childhood. In adulthood, the degree of anemia is usually mild, with an RC rarely greater than 6%. The anemia is usually normochromic and normocytic, and the overall prognosis is dependent on the underlying cause. In some cases, corticosteroids are helpful, and removal of the spleen, particularly during pregnancy, is used only as a last resort.


In microangiopathic hemolytic anemia (MHA), the products of destroyed red cells such as schistocytes are seen in the peripheral smear. Autoimmune diseases such as lupus erythematosus and polyarthritis have been associated with this disorder, as has pregnancy-induced hypertension, acute glomerular nephritis, and idiopathic thrombocytopenia purpura. In addition, other disorders associated with disseminated intravascular coagulation (DIC), such as fetal death, amniotic fluid embolus, and severe infection, may reveal these signs of MHA. In most of these diseases, small-vessel damage or abnormal coagulation leads to a network of fibrin strands that partially occlude the microcirculation. It is this basic pathophysiologic change that causes the fragmentation and lysis of the RBC. Therapy for those patients with MHA involves diagnosis and treatment of the underlying disorder and principally the reestablishment of the integrity of the microcirculation. The prognosis for both mother and fetus depends upon the alacrity with which the underlying disorder can be overcome.


In some patients, a positive direct Coombs test is found in conjunction with hemolytic anemia. In these patients, usually complement of the IgG antibody variety is irreversibly fixed to the erythrocyte membrane. This leads to premature destruction of the RBCs. These types of disorders usually are categorized as having “warm” or “cold” antibody hemolytic anemia. Those with warm antibody hemolytic anemias are usually associated with an IgG antibody and noted in many collagen-vascular diseases, patients with lymphomas, or drug reactions (particularly α-methyldopa). Those with cold autoantibodies are usually of the IgM class and are most commonly noted after a viral or mycoplasmal infection. The IgM antibody does not affix to the RBC membrane but rather stimulates complement, which is responsible for the positive antibody test. If the underlying disorder is treated or removed, the hemolytic process is usually assuaged in 2 to 3 weeks, although the Coombs test may remain positive for up to a year. Transfusion is not used unless absolutely necessary because the transfused cells likewise undergo fragmentation. Frequently, corticosteroids are used in those with warm antibody or IgG-mediated disorder. The prognosis for mother and fetus is dependent on the prognosis of the underlying disease process.

Intrinsic Hemolytic Anemia

Intrinsic hemolytic anemias are those in which there are inherited disorders leading to premature destruction of the RBCs. Included in these types of anemias are hereditary spherocytosis, various RBC enzyme deficiencies, and the hemoglobinopathies.24


This anemia is inherited as an autosomal dominant trait, and the RBC membrane is abnormally permeable to sodium. This leads to cellular disruption and also to premature cell death in the spleen. The diagnosis of hereditary spherocytosis is confirmed with the osmotic fragility test, and the treatment is usually transfusions or splenectomy. The prognosis for mother and fetus is usually excellent unless splenectomy must be performed during pregnancy. In most cases, treatment of this disorder with splenectomy has already been carried out prior to conception.


Various enzymatic defects inherited in RBCs can also lead to premature destruction of the erythrocyte. The two most common disorders involve pyruvate kinase deficiency and glucose-6-phosphate dehydrogenase (G6PD) deficiency. The pyruvate kinase deficiency is inherited as an autosomal recessive trait and is usually diagnosed during childhood because of persistent low-grade anemia and jaundice. The fluorescent spot test is the diagnostic test of choice. This is usually a mild disorder and does not disturb maternal or fetal homeostasis. G6PD deficiency, on the other hand, is inherited as an X-linked disorder and is more commonly found in blacks, with 13% of American black males being affected. The use of oxidizing drugs, such as nitrofurantoin and primaquine, stimulates acute hemolysis of red cells with accompanying hemoglobinuria. In addition, vital or certain bacterial infections may also stimulate this hemolytic event. The methemoglobin reduction test is the diagnostic assessment technique of choice, and treatment is usually removal of the offending agent. It is rare in today's society that this disorder would adversely affect the mother or developing fetus. Nevertheless, G6PD deficiency should be kept in mind prior to prescribing drugs to black parturients, since 25% of American black women are carriers of this trait.


Most of the disorders regarding hemoglobin usually involve the abnormal globin structure of the RBC. Fortunately, most of these abnormal hemoglobins, called hemoglobinopathies, are very rare; of the more than 300 types reported, only a few are clinically very common. Most of these disorders involve a single amino acid substitution in one of the globin chains, although some involve deletions, additions, or fusions in the α- or β-chains. All of these hemoglobinopathies, when clinically significant, lead to the production of erythrocytes that either are structurally unable to perform duties, such as oxygen transport, or are prematurely destroyed. All are associated with an increased reticulocytosis, indicating the body's ability to attempt compensation. To understand these, familiarity with hemoglobin synthesis and structure is necessary.

Erythrocyte Composition.

The erythrocyte is extremely complex; its membrane is composed of lipids and proteins, while the interior of the cell contains primarily hemoglobin, which is intimately associated with oxygen transport. The erythrocyte develops from the multipotential stem cell or myeloblast, which differentiates along the erythroid line in the bone marrow (see Fig. 2). During maturation, it loses most of the metabolic and biosynthetic capabilities inherent in most other cells. The capacity to synthesize nucleoproteins, such as DNA, is lost at the basophilic normoblast stage, and stainable DNA itself is removed before reticulocyte release, as shown in Fig. 2. Ribonucleic acid (RNA) synthesis is lost as the peripheral reticulocyte matures, and loss of electron transport mechanisms prevents the use of phosphate bonds as an energy vehicle, requiring the mature erythrocyte to employ anaerobic glycolysis by the hexose and penrose phosphate shunt pathways. Thus, the RBC's only metabolic function is the process of gaseous and ionic transfer, using the remaining pathways for purine and pyrimidine metabolism. Although these critical functions are performed well in the mature RBC, the lack of RNA and DNA is extremely important, since it renders the cell incapable of repair or repetition. Therefore, in certain disease processes that alter this maturation sequence (see Fig. 2), the production rate and functioning of the cell lines are impaired.

Hemoglobin Structure.

The structure of hemoglobin is shown in Fig. 10 and consists of heme, which is composed of a divalent iron atom connected to four pyrrole rings through their substituent nitrogen atoms.25 This polyvalent compound is surrounded by two pairs of dissimilar polypeptide chains, each linked to one of the pyrrole rings. Each pair of polypeptide chains has different sequences of amino acids and is stereochemically arranged in four helical configurations. The physical properties of heme, as well as its association with the iron atom, classify it as a metalloporphyrin, a structure similar to chlorophyll or vitamin B12. The conjugated double bonds of heme readily absorb visible light and are responsible for the red color of hemoglobin. The intense spectrophotometric absorption peak of hemoglobin at 412 nm, the Soret band, may be clinically important because it may overlap the peak of amniotic fluid bilirubin, measurement of which is used to assess the severity of Rh isoimmunization. Functionally, it is the iron atom that carries the oxygen; in its ferrous state (Fe2+ ), six binding sites are present. Four are attached to the heme, while another is attached to a histidine residue of a globin chain. The other coordination position of iron is unoccupied in deoxyhemoglobin under proper physiologic conditions. This final binding point is protected from oxidation by the globin chains surrounding the heme moiety. Separation of heme from the globin during destruction results in the oxidation of the iron atom and the formation of hematin. If hemoglobin iron is oxidized to its ferric state (Fe3+), methemoglobin is formed; this cannot serve as the oxygen carrier since the final coordination site is bound.

Fig. 10. Biosynthesis of hemoglobin.

Heme Synthesis.

Heme synthesis takes place in the particulate and soluble fraction of the erythrocyte (see Fig. 10). Initially, δ-aminolevulinic acid (ALA) is synthesized from succinyl-CoA and glycine in the particulate fraction of the cell.25 Next, two moles of ALA combine to form porphobilinogen, and subsequently, four of these moieties converge to form one mole of uroporphyrinogen III. Through several intermediate steps, this compound is converted into coproporphyrinogen III. By successive decarboxylations, it is converted into the heme precursor protoporphyrin IX, a reaction catalyzed by a mitochondrial enzyme called ferrochelatase or heme synthetase. A ferrous iron moiety is inserted into the protoporphyrin as a final step in the pathway of formation of heme. Current data suggest that heme biosynthesis is unidirectional and the control mechanisms for the pathway are necessarily located in the first enzymatic step (ALA synthesis). Positive feedback inhibition of the ALA pathway by certain metabolic byproducts such as hematin or hemin occurs, while other degradatory products such as billverdin and bilirubin do not appear to affect heme synthesis. Finally, the lack of precursors or cofactors, such as Fe2+, folic acid, vitamin B6, and vitamin B12, may impair heme production at various reaction sites during hemoglobin synthesis.

Globin Chain Synthesis.

The sequence of amino acids in the globin chains is determined by genetic control. Once initiated, the sequence of the nucleotide bases in DNA is transcribed to messenger RNA (mRNA). The mRNA is transported from the nucleus to the ribosomal RNA (rRNA), which is the actual site of protein synthesis. The mRNA, by collation with transfer RNA (tRNA) on the rRNA template, translates the message from a language of nucleotide bases into that of amino acids in the proper sequence. The initiation of the translation always begins with the same code word (AUG) from mRNA regardless of the globin chain synthesized; therefore, the first amino acid is always methionine. The initial phase, called transcription, proceeds through approximately 40 amino acid subunits. The elongation phase extends nearly to the termination point and is catalyzed by a protein called elongation factor I, which is glucose triphosphate (GTP) and tRNA dependent. Finally, the parent mRNA gives the signal to terminate transcription and the newly formed globin chain is released, while the ribosome searches for the initial phase of another mRNA. The three nucleotide codons that signal for the chain termination of all proteins are UAA, UAG, and UGA. These are decoded by tRNA and by protein re leasing factors in the particulate cell fraction. Once complete, each polypeptide chain contains 141 to 146 amino acids. After synthesis, two pair of globin chains interact with the four pyrrole portions of the heme and the assembly of the hemoglobin moiety is complete (see Fig. 10).

Hemoglobin Nomenclature.

The polypeptide chains in normal adult hemoglobin are termed alpha (α), beta (β), gamma (γ), and delta (δ). The nomenclature is based on various alterations in the globin chain and is summarized in Table 8. The α-chains are found in all normal adult hemoglobin and contain 141 amino acids. They combine with the β-, γ-, or δ-chains, each containing 146 residues, to form the two pairs of globin chains found in normal hemoglobin. For instance, when two β-chains accompany two α-chains (α2A β2A ) hemoglobin A (HbA) is formed; this accounts for 95% of the total hemoglobin in the normal adult. Hemoglobin A2 (HbA2) is formed by two δ-chains accompanying the α-chains (α2A δ2A) and accounts for 2% to 3.5% of normal hemoglobin. Fetal hemoglobin (HbF) is formed by two γ-chains coupled with two α-chains (α2A γ2F) and accounts for the remainder of hemoglobin found in the adult.

TABLE 8. Nomenclature of Normal and Abnormal Hemoglobins









Common Name

Official Designation

Adult hemoglobin*








Adult hemoglobin*








Adult fetal hemoglobin*








Embryonic fetal hemoglobin*


HbGower II






Sickle hemoglobin†



α2Aβ2S(6 gluval)








α2Aβ2C(6 glulys)








α2M(58 his tyr)β2A





Increased oxygen



α2Ches(92 argleu)β2A





Unstable hemoglobin†



α2Aβ2Saki(14 leupro)






None (decreased







* Structurally normal variants.
† Structural abnormalities include substitution, inversion, or deletions of amino acids in either chain.
‡ Other abnormalities may affect either chain.

An amino acid deletion, inversion, or substitution structurally alters the globin chains to cause most of the abnormal hemoglobins. The location of an amino acid abnormality (if known) is designated by a superscript notation (in parentheses)next to the involved globin chain: α2A β2S(6 gluval).26 Abnormal hemoglobins may also involve the α-chains and are called α-hemoglobinopathies. If several abnormal hemoglobins are similar in their biochemical properties, a subscript is used to indicate the geographic location where the hemoglobinopathy was first described (e.g., HbMBoston, HbMMilwaukee). The changes in the amino acid sequence of the globin chains may evoke a wide spectrum of clinical symptoms. Abnormal hemoglobins may also be classified according to the changes they elicit in oxygen carrying capacity, stability, erythrocyte longevity, or synthesis rate (see Table 8).

Hemoglobin Genetics.

The structure and function of the globin chains are determined by polygenetic inheritance, whereas there appears to be genetic control for heme synthesis common to all hemopoietic tissue. The four globin chains contain hundreds of amino acids as possible mutagenic sites, and since there may be at least four mutations per position by nucleotide sequencing, the number of changes in hemoglobin structure by genetic malfunction is enormous. Four separate genetic loci have been identified for the α, β, γ, and δ polypeptide chains. The use of marker chromosomes in cytogenetic studies indicates that the genetic loci for the α-chain is on chromosome 16, while the β-, γ-, δ-chain loci are closely associated on chromosome 11. The structural changes in the amino acid sequencing characterizing most abnormal hemoglobins appear to be related to genetic mutation. In contrast, thalassemia seems to occur as a result of defective mRNA. The cause of the various mutations responsible for the globin chain sequence for most hemoglobinopathies has not been determined; however, the mutation in hemoglobin S (HbS) may be a genetic response to malaria. Due to clinical symptomatology in various hemoglobinopathies, there is no clear dominant or recessive inheritance pattern; thus, the mode of inheritance for most hemoglobinopathies is autosomal codominant.

Sickle Hemoglobinopathies.

Sickle hemoglobin is rather common in this country among blacks. HbS-S (sickle cell anemia) and its more severe variants, HbS-C and HbS-Thal, comprise a group of hemoglobinopathies called sickle cell disease (SCD).27 It is these patients who are most severely affected clinically during pregnancy or in the nonpregnant state. Those affected with sickle cell trait and mild sickle cell anemia (HbS-Memphis, HbS-D) may be asymptomatic during gestation.

Hemoglobin S represents a structural defect caused by genetic mutation, possibly as an adaptive response to falciparum malaria. Heterozygosity (HbA-S) gives a protective effect, while those persons who are HbA-A or HbS-S die during childhood in areas endemic for malaria. The substitution of the amino acid valine for glutamic acid at the sixth position from the N-terminal of both β-chains gives S hemoglobin its unique sickling characteristic compared with normal adult HbA. Likewise, the substitution of lysine at the identical position results in hemoglobin C (HbC). Otherwise, these hemoglobins are exactly the same as normal adult hemoglobin, and this one change is responsible for all of the symptomatology. The amino acid substitution also causes the differences in electrophoretic migration of HbS and HbC, HbA, HbF, and HbA2. Changes caused by amino acid substitutions, deletions, or inversions in either the α- or β-chain give rise to other structurally abnormal variants, and although some of these hemoglobins have an adverse affect on pregnancy outcome, they are fortunately quite rare. Therefore, hemoglobinopathies listed in the SCD category are the only ones that are clinically important. Mild forms of the disease are important only when the patient is under severe stress or for genetic counseling purposes, since the clinical presentation is usually benign.


The terminology of sickle hemoglobinopathies has undergone many revisions since 1960. The currently accepted classifications include (1) sickling disorder, any red cell that undergoes in vitro/in vivo sickling in the deoxygenated state, (2) sickle cell trait (HbA-S), the heterozygous form of HbS infrequently associated with clinical symptoms, (3) sickle cell anemia (HbS-S), the homozygous form of HbS frequently associated with severe clinical symptoms, (4) sickle cell disease (HbS-S, HbS-C, and HbS-Thal), disorders in which HbS is all or part of the abnormal hemoglobin composition and that are usually associated with severe symptoms, and (5) mild sickle cell states (HbS-D, HbS-E, HbS-Memphis), those in which HbS is present but clinical symptomatology is usually mild. Numerically, HbS, both homozygous and in combination with other hemoglobins, is the most common abnormal structural defect; HbC is the second most common hemoglobinopathy in this country, and HbE is next in frequency. Hemoglobin D is similar in its other properties to HbS but does not sickle. Both HbE and HbD can combine with HbS to give mild clinical symptomatology.


Hemoglobin functions quite well in the oxyhemoglobin form with respect to oxygen affinity, heme interaction, Bohr effect, and reactivity with 2,3-diphosphoglycerate (2,3-DPG). In the deoxygenated state, the abnormal valine forms hydrophobic bonds with adjacent amino acids. This aggregation results in the formation of tetramers, which are relatively insoluble and coalesce to form microcables. These structures “gel” with similar aggregates to form a cable that distorts the cell into the classic sickled erythrocyte. As this process continues, the cell membrane is affected by loss of cations and phosphorylation capability; thus, the cell becomes irreversibly sickled. The age of the cell, concentration of S and F hemoglobin, extent of oxygenation, status of the cell membrane, temperature, pH, and osmotic pressure of the milieu in the microcirculation all play a role in determining whether a given cell will sickle. The initiation of the crisis may be the formation of sickled cells, with further deoxygenation in the microvasculature, creating a vicious cycle (Fig. 11). Continued sequestration of the cells may lead to the acute and chronic morphologic changes found in the organs of patients with sickle cell anemia. Clinically, this sequence of events causes the pain, leg ulcers, hepatomegaly, autosplenectomy, and chronic anemia found in these subjects. As a rule, hemoglobinopathies not associated with HbS, unless homozygous, usually present: fewer management problems to the clinician than do those in the SCD category. Many of these (HbA-C, HbC-C, HbA-D, heterozygous thalassemia) are usually diagnosed only during the workup of a patient with mild, iron-unresponsive anemia.

Fig. 11. The sickle cell crisis cycle.(Martin JN Jr, Morrison JC: Managing the parturient with sickle cell crisis. In Huddleston JF [ed]: Clinical Obstetrics and Gynecology: Sickle Cell in the Gravida Woman. Philadelphia, JB Lippincott, 1984)

Laboratory Assessment.

Anemia is the most constant feature of subjects with severe hemoglobinopathy. Generally, patients with consistent PCV less than 20% or a hemoglobin level below 6 g/dL have homozygous β-thalassemia or HbS-S. Patients with PCV of 20% to 25% may have a HbS-S, HbS-C, or HbS-Thal hemoglobinopathy, while subjects with a PCV of 25% to 30% may have any hemoglobinopathy except homozygous β-thalassemia and HbS-S. In most of the mild forms (HbS-D, HbS-E, HbA-S), the PCV is 30% to 35%. When the PCV is 25% to 35%, diagnosis is particularly difficult because many disorders are included in the differential. The RBCs are usually normochromic, normocytic, but concomitant IDA may confuse the issue. A reticulocytosis of up to 20% and thrombocytosis are not uncommon in HbS-S patients during crisis. The serum bilirubin is usually mildly elevated in the asymptomatic patient with HbS-S but may reach 15% to 20% during crisis. The urinalysis usually shows hyposthenuria (with specific gravity <1.005), hematuria, and bile and direct bilirubin. Asymptomatic bacteriuria is more common in these patients, and pyuria usually leads to active symptomatic infection due to poor opsonization of bacteria. The liver enzymes are usually elevated and reach high levels during crisis; abnormalities of coagulation, such as a decrease in the platelet count and fibrinogen with presence of fibrin split products, are more common during a crisis but may be present at any time. The prothrombin, partial thromboplastin, and clotting times are normal in patients with sickle hemoglobinopathies.

Several commercial kits are available for diagnosis of HbS (Sickledex, Ortho Diagnostics, for example) and are based on turbidimetric assessment or color change.28 Although these tests have few false negative reactions, they do have a 5% to 8% incidence of false positive reactions; the standard deoxygenation assay using a reducing agent (sodium metabisulfate) is employed on positive screening samples. This microscopic test is very accurate but does not differentiate the various hemoglobinopathies. Electrophoresis and chromatography with starch gel, cellulose acetate, and agar gel are used to quantitate each hemoglobin and to aid in the diagnosis of the specific hemoglobinopathy. By electrophoresis, HbA-S usually has 25% to 35% HbS, 60% HbA, normal HbA2, and increased HbF; HbS-S demonstrates 90% to 95% HbS, with small concentrations of HbA2 and normal to increased amounts of HbF; HbS-C usually evidences normal percentages of HbA2 and HbF, with an equal division of the remainder between HbS and HbC (45% to 48%).29 Hemoglobin S-Thal is very difficult to diagnose even by electrophoresis but appears to be much like HbA-S with slightly more sickle hemoglobin.

Clinical Presentation.

Hemoglobin S-S, HbS-C, and HbS-Thal can be usually differentiated from the benign hemoglobinopathies, such as HbA-S, by a history of recurrent painful vaso-occlusive crises in the former. These crises usually last 2 to 6 days, may be associated with fever as well as leukocytosis, and may lead to unnecessary surgical procedures unless the hematologic diagnosis is suspected. These attacks are more common in HbS-S patients, usually average one to four a year, and are commonly associated with fever, pregnancy, trauma, or physical stress. The crisis is rarely truly “hemolytic” and is more commonly termed vaso occlusive, although other types of crises are possible (see Fig. 11). Finally, cerebral manifestations are not extremely common but are ominous, since they may be associated with severe morbidity and death. The sudden deaths with SCD have been reported to involve both coronary and cerebral vessel occlusions. Coma, convulsions, and death due to cerebral infarction are most common in these patients. The sudden onset of symptoms after a benign history is characteristic of patients with HbS-C. Up to 20% of patients with severe sickle hemoglobinopathies die by age 10, and many more die by the age of 40.

Sickle Cell Disease in Pregnancy.

Incidence. The hemoglobinopathies comprising SCD are more common than they appeared to be in the past. Hemoglobin S-S was said to occur in 1 in 600 Americans of African descent; however, the incidence is closer to 0.5% to 0.7% in certain areas in this country. Hemoglobin S-C is present in about 1 in 800 persons of African descent.27 The incidence of HbS-Thal is said to be 1 in 1250, but this is probably underreported, since the anemia is mild. The fertility of patients with SCD was assumed to be reduced because there are few pregnancies, but newer evidence shows that the incidence of pregnancy in these subjects is the same as in the population at large.

Maternal Effects.

The prevailing attitude of most perinatologists is that the patient with SCD appears to be at significantly greater risk for crisis and other dire complications when pregnant.29 Only infection, trauma, hypoxia, and acidosis approach pregnancy as a frequent cause of crisis and severe morbidity/mortality in patients with SCD. Factors associated with pregnancy that complicate the clinical picture include the “hypercoagulable state” of pregnancy, the increased susceptibility to infections, the increased vascular stasis in the pelvis and lower extremities, the increased metabolic and hematologic demands of pregnancy, and the stress of parturition. Because of these factors, the status of mother and fetus was very poor in many series.

Several large review series prior to 1969 show morbidity rate of 50% to 80% and maternal mortality figures ranging from 2% to 25% for SCD patients during pregnancy.29,30 Painful crises were among the most common symptoms, being reported by all investigators. In addition, most patients had a reduction in their hematologic values during gestation. Other major complications such as pneumonia (3% to 15%), pyelonephritis (5% to 12%), and endometritis (72% to 10%) are frequent causes of crisis. Pulmonary embolus (1% to 9%) and congestive heart failure (1% to 5%) are also common causes of severe morbidity, as is pregnancy induced hypertension. Since 1970, the maternal death rate has dropped considerably (i.e., 0 to 1% in most large series), as has the morbidity rate in this country. These changes probably reflect better antepartal scrutiny as well as new management techniques.

Although some authors feel that the clinical course of SCD in pregnancy tends to be similar to its course before pregnancy, many previously asymptomatic patients may suffer a serious, sometimes fatal, crisis during pregnancy.30 Also, several investigators have associated an increased level of hemoglobin F with a reduction in the number of complications. The hypothesis is supported by those rare persons with persistent hereditary fetal hemoglobin who have no crises. However, with persistent hereditary fetal hemoglobin, the hemoglobin F is homogeneously distributed in each cell, whereas in patients with SCD, the hemoglobin F, although it may be present in large quantities, is unequally distributed in the cell. Several studies have shown that the level of hemoglobin F varies greatly in both asymptomatic patients and those in crisis. Therefore, it is unwise to rely on a negative history or absolute hemoglobin F percentage to guide therapeutic judgment in pregnant patients with SCD.

Although hemoglobin is more stable than hemoglobin S, the combination of hemoglobin C with hemoglobin S appears to facilitate sickling more than does a combination containing hemoglobin A. This is particularly true in late pregnancy, when venous stysis and maternal hypercarbia are usually present. It has been shown that the number of sickled cells remains constant for a given Po2 if hemoglobin S is present alone. However, if both hemoglobin C and hemoglobin S are present, the number of distorted cells increases to a greater degree under similar conditions. The red cells appear to have a shorter survival time in HbS-C patients due to the crystallization of hemoglobin C within the red cell. Symptomatically, patients with HbS-C appear to be healthier and have fewer crises than do those with HbS-S, but complications resulting from stress, such as pregnancy, infection, or trauma, are more severe and sudden (e.g., serious eye lesions, renal papillary necrosis, aseptic bone necrosis, acute pulmonary thrombosis, marrow emboli, and true hemolytic episodes). In general, pregnancy results in a similar risk for patients with HbS-C and HbS-S with regard to morbidity and mortality, while patients with HbS-Thal have a clinical course intermediate between those with HbS-S and HbA-S.

Fetal Effects.

Perinatal salvage rates in patients with SCD have been very low in most series. The incidence of abortion is high (20% to 35%) and has been related to the “chronic disease state” or malnourished condition of the patient, sickling in the infundibulopelvic vessels (decreased progesterone), and distorted cells in the arcuate arteries of the uterus (decreased placental blood flow). These changes may also lead to intrauterine growth retardation and premature labor, which is also increased (10% to 55%) in gravid patients with SCD.29,30 The decrease in mean birth weight of these infants (250 to 500 g) has also been related to the high incidence of maternal infection and premature labor. Stillbirths are also increased (5% to 13%) and appear to be related to severe maternal crises with evidence of placental sickling. Although fetal deaths occur in asymptomatic subjects, there have been no reported stillbirths due to sickling in the HbS-S fetus. The overall neonatal mortality rates are not increased for progeny of women with SCD, provided asphyxia and growth related complications do not occur; actually, the incidence and severity of respiratory distress syndrome are decreased in these offspring.30

The perinatal salvage rates are decreased in patients with HbS-C, ranging from 40% to 85%. There is an increase in abortions and fetal deaths. Both prematurity and small-for-gestational-age infants are more common, but not to the extent found in HbS-S patients. Hemoglobin S-Thal patients show even a further reduction in abortions and fetal deaths compared with HbS-S and HbS-C patients. The percentage of small-for-gestational-age and premature infants is increased in HbS-Thal and near average for HbS-C patients.

Antepartum Management.

Intensive antepartal management of SCD patients, with visits at bimonthly intervals for the first 20 weeks and then weekly until delivery, appears to be the best method of preventing decompensation and crisis. During each visit a thorough search should be made for infection and historical and physical evidence of sickling; also, a nutritional survey should be made and diet counseling should be provided. Vigilance for crisis is important, and the patient should be hospitalized if pain, fever, infection, or other signs of sickling are noted. PCV and hemoglobin determinations should be performed on each visit, since rapid reductions in these indices may occur even in the asymptomatic patient. The RC, blood film, and serum bilirubin determination may also be important in assessing the patient's hematologic risk for crisis. Methods for assessment of fetal well being, such as ultrasonography, oxytocin challenge test, nonstress test, and biophysical profile, are also used to detect fetal jeopardy in these patients.

Folic acid, 0.5 mg twice daily, is recommended in pregnant SCD patients. Iron therapy, in contrast, was thought to be ineffective and etiologically related to hemochromatosis if administered continually. However, IDA is common in these patients and was found in as many as 65% of pregnant HbS-S patients.31 Because the duration of pregnancy is short, many investigators administer 60 mg of elemental iron three times a day with meals during pregnancy if the blood smear or other hematologic assessment techniques show evidence of IDA in patients with SCD.

Therapy of patients with SCD has been largely symptomatic. Analgesics, hydration, vasodilatation, oxygen, bed rest, systemic alkalinization, and anticoagulation have all been used with questionable success. Tolazine, cobaltous chloride, carbon monoxide, and carbonic anhydrase inhibitors have not been found useful. Hyperbaric oxygen, although effective, is cumbersome and most often unavailable. In general, only oxygen and mild analgesics are still used clinically in the treatment of crisis. Desickling agents, such as urea and cyanate compounds, prevent deoxyhemoglobin S from sickling and increase the oxy configuration of the hemoglobin molecule to actually prolong the RBC life span. Unfortunately, to prevent or reverse sickling, urea must be administered in high doses to reach a serum concentration of 150 to 200 mg/dL. The effect of this high urea concentration, as well as the posttreatment diuresis, on the parturient and fetus is not known; thus, this drug is not used during pregnancy. Likewise, potassium cyanate and carbamylphosphate therapy have been found to reverse sickling in vitro, but their use is complicated by peripheral neuropathies and therefore is curtailed.

More recently, there have been developments of other compounds such as adipimidates and the use of various alcohols. Although these are still too toxic to be employed in humans, promising research indicates that breakthroughs in the near future are almost certain. On the other hand, other desickling substances such as peptides, particularly those containing tryptophan and tyrosine, appear to be completely nontoxic.32 Promising research in these areas has been hindered only by the fact that large quantities are necessary to metabolically desickle the sickled hemoglobin. Therefore, although these peptides work nicely in vitro, large enough quantities cannot be infused at the present time to make it clinically worthwhile. Reduction and concentration techniques are currently underway with these compounds. Also, genetic manipulation by induction of hemoglobin A or hemoglobin F has been successful in higher mammals. Genes for normal hemoglobins have been inserted in subhuman primates and appear, at least for short periods of time, to be the producers of normal hemoglobin. Longevity studies and infectious complications, however, await further modification before this can be extrapolated to humans. The use of recombinant human erythropoietin increases fetal hemoglobin production but the levels achieved are not sufficient to curtail symptoms of sickling, making this modality of limited value at present.23

The use of induced hyponatremia is also a very worthwhile technique, with the only problem being the dangers of hyponatremia itself. Since the sodium must be reduced to very low levels and the inherent risk for convulsions is so great, this treatment is for those in severe intractable crises. Finally, the treatment of the patient with sickle hemoglobinopathies with drugs that alter hydrophobic bonding appears very promising. The use of the drug Cetiedil in African countries and in Europe may be our first good method of prophylaxis. In early trials, the patients with severe sickling history appear to be asymptomatic with the use of this drug. Furthermore, this drug has been used for approximately 20 years in the nonpregnant and pregnant state as a local anesthetic with no hint of any disorder in the mother or fetus. We await further trials with this agent. In summary, the development of new treatments awaits further testing of normal hemoglobin induction and of the further concentrating efforts of the peptide agents. In addition, work with Cetiedil for oral prophylaxis also appears promising.

Another therapeutic approach, the partial exchange transfusion, began to be used in 1962, was advocated for pregnancy in 1965, and has enjoyed moderate success since 1970.34,35,36 While most authors recommended transfusions for severely anemic SCD patients (with hemoglobin levels <5 g/dL), some give blood prophylactically.34,36 The use of this technique in a high risk parturient may be desirable, particularly with sophisticated blood bank assistance. The principal advantages of this technique are several: (1) it dilutes the concentration of hemoglobin S—containing cells with normal erythrocytes containing hemoglobin A, thus favoring oxyhemoglobin formation while reducing stasis and subsequent sickling, (2) the clotting factors that are increased during pregnancy may be diminished, while the production of the sickle erythrocyte is halted, as evidenced by the low RC, and (3) it can be used in the presence of renal or cardiovascular failure when simple transfusions may not be well tolerated. The disadvantages of this modality are those associated with blood administration itself. These include febrile reactions, premature labor, development of isosensitization, iron overload, HIV, and hepatitis. With meticulous screening of donor blood, the use of buffy-coat—poor washed red cells, volunteer blood donor (family member or friend), and blood that has been assessed for venereal diseases, as well as various hemoglobinopathies, the above problems are less likely to occur. It should be noted, however, that bizarre clinical manifestations may occur with these transfusions. Substernal chest pain has been noted shortly after or during transfusions; the cardiovascular and hematologic assessments were normal in these patients, but the etiology remains questionable. Also, the incidence of blood antibody development has not been fully studied. For this reason, it is recommended that these transfusion protocols be limited to level III institutions where adequate hematologic and obstetric personnel, as well as facilities, are available to cope with these and other problems that may be engendered by transfusion.

The method of transfusion has caused great controversy. The transfusion may be done manually: the patient's blood is phlebotomized and replaced with donor red cells. In a more modern technique, erythrocytapheresis, automated blood separator devices are used for this purpose. Because the IBM Cell Separator* is continuous, it is the preferred technique. The principal advantages of erythrocytapheresis for these patients are that the procedure can be performed in 1 to 2 hours and on an ambulatory basis, thus reducing that expense to the patient, even though erythrocytapheresis is relatively expensive. In addition, careful control over the amount of RBCs removed and infused is afforded the patient. Finally, there is an advantage in that the patient's own plasma and platelets and granulocytes can be returned rather than using crystalloid solutions as the diluent. The use of erythrocytapheresis with the IBM Cell Separator is currently the method of choice for those who perform exchange transfusions in sickle hemoglobinopathy patients during pregnancy.37

If transfusions are used, some investigators begin at 24 to 28 weeks, since 95% of the serious complications occur in the third trimester. Others have advocated initiation of therapy as soon as pregnancy is identified, to promote the growth and development of the fetus and to ameliorate those cases in which infection or crisis might occur earlier.38 A PCV of 30% to 35% with 40% hemoglobin A appears to be a reasonable goal, although with such sophisticated exchange equipment or hypertransfusion, 90% to 95% hemoglobin A is possible. The beneficial effect usually lasts 6 to 8 weeks, and the transfusions are repeated if the PVC falls below 25%, the hemoglobin A falls below 20%, a crisis occurs, or labor ensues. Other investigators prefer to use exchange transfusions until the hemoglobin A is at least 60% to 70% and advocate retransfusions if the hemoglobin A falls below 50%. Usually, higher levels of hemoglobin A have not been necessary to prevent the untoward effects of sickle hemoglobinopathy on the reproductive outcome. If patients are admitted in labor and the hemoglobin A is not acceptable, transfusions may be performed at that time.

Studies using transfusions and intensive antepartal risk assessment have demonstrated a remarkable decrease in maternal and perinatal morbidity and mortality regardless of when the transfusions are started34,36 The use of prophylactic exchange transfusions clearly decreases the incidence of vaso occlusive crisis compared to conservative management, while perinatal outcome appears to be equally good with both techniques.39,40

It must be stated that some investigators have obtained equally good results by supportive therapy without transfusions.41,42 Although the number of patients in each series is small, frequent antepartal scrutiny combined with aggressive management of crises and infection seems to yield acceptable results. Therefore, it is apparent that there is controversy over whether to use transfusion or not, as well as what type of transfusion (hypertransfusion, exchange, simple, prophylactic) to employ, if this method is to be used. In my experience, prophylactic transfusions often have yielded good results and appear to be essential to obtaining a good reproductive outcome in these patients. Clearly, a prospective randomized study is needed to answer the question of which antepartal therapy is best for the pregnant SCD patient.

Intrapartum Management.

During labor and delivery, segmental epidural block appears to be the best method of analgesia if performed by the obstetric anesthesiologist or skilled perinatologist; hypotension and resultant hypoxia greatly increase the risk of sickling. Alternatively, small amounts of meperidine or other analgesic agents may be used with pudendal or spinal anesthesia for delivery. Oxygen is administered at 4 L/minute with the patient laboring in the semi-Fowler, lateral recumbent position. Patients may be followed by blood smears to detect: sickling, and the fetus is assessed by continuous electronic monitoring as well as scalp blood sampling when necessary. Delivery is usually spontaneous at term, and cesarean sections are performed only for obstetric indications. Oxytocin has not been found harmful in the patient with SCD and is used when indicated.

Postpartum Management.

Postpartum SCD patients are scrutinized for excess blood loss, infection, or thrombophlebitis. Antithrombotic stockings, if properly applied, and early ambulation are encouraged; prophylactic antibiotics are usually not employed. Frequent visits (biweekly) to the clinic assure normality through the 12-week postpartum adjustment phase. The infant should be tested using cord blood for electrophoresis to detect heterozygotes and to identify those with SCD so that counseling and education of the parents can effect a more favorable outlook for the child.

In patients with SCD, the use of oral contraceptives is contraindicated since thrombosis, tissue infarction, and sickling may be enhanced by the estrogen component of these agents. In contrast, data from developing countries show large numbers of patients using progestins without complications. Further investigations using controlled trials must be performed, however, prior to recommendation of birth control pills for SCD patients in this country. Intrauterine devices are also not recommended in patients with SCD owing to alterations in the complement and properdin pathway that appear to be associated with an increased incidence of endometritis and tubo-ovarian abscesses. Medicated intrauterine devices (with copper and progesterone) have been used, and they appear to be more efficacious and safer in these patients, but they have not been thoroughly tested. Medroxyprogesterone (Depo-Provera) has been recommended for contraception in patients with SCD, since this agent may stabilize the sickle cell and actually reverse the sickling process in larger doses. Although the in vivo concentration is not large enough to prevent sickling, no harmful effect of the agent has been demonstrated and it appears to be an effective method of contraception in these patients. At this time, however, it is not available as a contraceptive agent. The barrier contraceptives such as condoms, diaphragms, and foams are extremely safe but are not very effective; therefore, if they are to be used, intensive patient education is mandatory to ensure compliance.

Much has been written about patients with severe hemoglobinopathies concerning surgical sterilization or abortion. Based on earlier data, some have recommended sterilization prior to pregnancy, while others have urged abortion and sterilization or sterilization after one pregnancy. More recent statistics, however, indicate that the danger to the patient from pregnancy is less than in earlier years. For this reason, the patient should be able to make a decision based on her own desires as to family size, taking into account the genetic disability of the infant, the risk of the therapy during pregnancy, and her potentially decreased longevity. Although all the maternal and fetal statistics are improved, it is still much more dangerous for SCD patients to endure pregnancy than it is for those subjects without a hemoglobinopathy. Patients with hemoglobinopathies not involving hemoglobin S can use any contraceptive prescribed for “normal” patients; no ill effects have been reported.

Sickle Cell Trait (HbA-S).

Sickle cell trait is a common hematologic abnormality and occurs in 5% to 14% of blacks in this country. The patient is usually not anemic unless a concomitant IDA is present. The benign clinical course and inadequate screening lead to underdiagnosis of the HbA-S. The fertility of these patients is thought to be normal. Sickle trait has been associated with an increased susceptibility to renal problems, such as hematuria, frequent infections, and hyposthenuria. Splenic infarction has been reported, but actual crisis is rare unless hypoxia or infection is present, Studies in the literature concerning pregnant patients with HbA-S are less well documented and less numerous than those involving SCD. These patients appear to have an increased incidence of pyelonephritis, asymptomatic bacteriuria, and chronic renal disease when studied during pregnancy. The renal problems may be related to sickled cells lodging in the hypertonic portions of the renal medulla, leading to stasis, ischemia, and structural damage, followed by frequent infections. The persistent hyposthenuria may involve a similar mechanism, but these theories are unproved. Pneumonia and toxemia may be more common in pregnant HbA-S patients, but most data do not support this contention. Therefore, the maternal morbidity is much less than in patients with SCD and near that for similarly matched patients with HbA-A. The effect of sickle trait on the neonate is minimal, with most studies demonstrating that the mean birth weight, Apgar scores, prematurity rate, and abortions are identical to a matched control population of the hemoglobin A subjects.43

Antepartum management of these patients should be directed toward screening programs for detection, close scrutiny for positive urine cultures, and avoidance of intrapartum complications such as hypotension or blood loss, which may lead to hypoxia. Routinely, iron and vitamin supplementation are given to these women as to the normal parturient, since nutritional IDA may exist. If asymptomatic bacteriuria is present and persists after treatment or if a single episode of cystitis or pyelonephritis occurs, continuous antibiotic administration is recommended during the entire pregnancy.

Other Sickling Hemoglobinopathies.

Hereditary persistence of fetal hemoglobin (HPFH) occurs in 0.1% of blacks in the United States. The inheritance pattern for HPFH is also autosomal codominant, with allelic genes for hemoglobin A, S, and C. Whether pregnant or nonpregnant, persons with HPFH and hemoglobin S have a benign clinical course despite high concentrations of sickle hemoglobin, Persons with HPFH should be distinguished from other subjects with SCD who may have a high hemoglobin F. In HPFH-HbS the electrophoretic pattern is hemoglobin F, 25% to 30%; hemoglobin S, 70% to 75%; and hemoglobin A2, 2% to 3%. Persons with HbS-S rarely have over 20% hemoglobin F. More importantly, the hemoglobin F is heterogeneously distributed in routine hemoglobin S patients, with some erythrocytes having no hemoglobin F and others having 30% to 40%. HPFH patients, in contrast, have the hemoglobin F evenly distributed (approximately 30%) in each RBC, accounting for the lack of clinical symptoms. Hemoglobin S patients with HPFH should be reassured concerning their longevity and the benign clinical course.

Hemoglobin Lepore-S and hemoglobin O Arab-S are caused by a structural defect in the β-chain. They resemble HbS-S or HbS-C in symptomatology but their laboratory assessment mimics homozygous β-thalassemia. The therapy during pregnancy is similar to that for SCD because of the clinical severity.

Structural Hemoglobinopathies without Hemoglobin S.

Other clinically significant hemoglobin variants include unstable hemoglobins, the M or cyanotic hemoglobins, as well as those with abnormal oxygen affinity.44 Clinical management is dependent on the clinical manifestations as well as the severity of the disease.

Diagnosis of the unstable hemoglobins can be assisted by analysis of a peripheral blood smear. Normocytic, hypochromic cells with some poikilocytosis and basophilic stippling are seen. Inclusion bodies may be difficult to distinguish prior to splenectomy but appear as finely distributed dots in their early stages. Characteristically, the RC and serum iron values are high. Electrophoretic separation can confirm the diagnosis. When the degree of hemolysis is severe and an acute hemolytic crisis occurs, transfusions and splenectomy may be performed.

Congenital methemoglobinemia, or hemoglobin M, can be linked to two etiologies. A single amino acid substitution in one of the globin chains of hemoglobin is the more common cause. Five varieties of this abnormal hemoglobin have been identified. Substitutions in the α-chain have been identified in two varieties leading to cyanosis in infants at birth. The remaining three variants involve substitutions in the β-chains; these induce cyanosis by the third or fourth month of life. However, chemically induced or congenital deficiency of the enzyme methemoglobin reductase has also been linked to this disorder. Normally, oxygen dissociates from the hemoglobin molecule, leaving the iron with the hemoglobin in the ferrous state. When oxidation to the ferric form results, methemoglobin is produced and an ineffective molecule for oxygen binding results. When the level of met hemoglobin exceeds 2 g/dL, a state of cyanosis occurs.

Methemoglobinemia, resulting from amino acid substitution in a globin chain, has autosomal dominant transmission, whereas methemoglobin reductase deficiency is the result of autosomal recessive error. Although those affected are cyanotic, they remain asymptomatic. Diagnosis of hemoglobin M can be made by spectroscopy; the abnormal hemoglobins should first be identified by electrophoresis. More specific differentiation of the normal and abnormal hemoglobins can then be made using electron paramagnetic resonance to identify the amino acid substitution. Those affected with met hemoglobin reductase deficiency reveal normal spectral analysis. When hemolysis occurs, hemoglobin M can induce slight jaundice. As with the other hemolytic anemias, sulfonamides can exacerbate the hemolysis. Methylene blue and ascorbic acid have been useful in treating those with the enzyme deficiency; however, all modes of therapy are ineffective in treating those with the structural defect. Because the disease is benign and lacks symptomatology, pregnancy is not affected by these abnormal hemoglobins.

Abnormalities of oxygen affinity have been reported, with almost 50 variants already identified. High-affinity disorders account for three fourths of the variations, with 32 having been identified. Because of the increased affinity of the hemoglobin molecule to oxygen, smaller amounts of oxygen are released to the tissues. A compensatory polycythemia results in response to increased levels of erythropoietin. These variants are transmitted by autosomal dominant means and have been observed only in heterozygotes, since homozygosity is incompatible with life. A combination of starch-gel and agar-gel electrophoresis can identify the majority of these cases; however, oxygen affinity analysis can also confirm the diagnosis. Erythropoietin levels may also be useful in identifying these disorders. Cautious phlebotomy can be used to treat those symptomatic patients, bearing in mind that the polycythemia is a compensatory response that may be stimulated by a reduced hematocrit.

Overall life expectancy is unchanged for these patients; however, the incidence of fetal loss is increased. The pathophysiology of this disorder is the same as hemoglobin M. The etiology can indeed be associated with the structural variation and amino acid substitution but is also associated with a resultant decrease in levels of erythropoietin. Mild cyanosis seems to be the most prevalent symptom. Hemoglobin electrophoresis and electron paramagnetic resonance are the most useful diagnostic evaluations available.

The low-affinity disorders have no effect on pregnancy nor does pregnancy exacerbate this disorder. The clinical effect of these disorders is benign, and no treatment is indicated.

Genetic Counseling and Patient Education for Inherited Anemias.

Identification of patients with hemoglobinopathies and their subsequent education as to the effect of this disease process are important in obtaining optimal medical results. Mass screening programs for detection of hemoglobinopathies are available but have not been widely used because the public has a poor understanding of the heterozygous state. Many blacks with hemoglobin S detected by earlier programs were compromised economically when employers were told of their disability. At this time, screening for hemoglobin may be offered to blacks as they reach their teen years, on premarital examinations, or in various health clinics (obstetric, medical, and surgical). It is important, however, that such testing be offered but not required. Many states require testing for sickle hemoglobin; however, the compulsory nature of such laws has made much of the testing unacceptable to many blacks.

Another major issue is that of counseling the pregnant woman with hemoglobin S concerning the transmission of a hemoglobinopathy to her offspring. If the father is also known to be a carrier of hemoglobin S, there is a 25% chance that the child will have hemoglobin S, a 25% chance that the child will have hemoglobin A, and a 50% chance that the child will inherit the trait. On the other hand, if the hemoglobin of the father is not known, one can say that the odds are about 2% that the child will have HbS-S, given an 8% incidence of sickle cell trait among blacks in this country. This can be extrapolated to a 4% chance (1:25) that the infant will have HbA-S. In patients with HbS-S, the question is more difficult, since if the father is HbA-S, 50% of the infants will have HbS-S and 50% will receive the trait. Therefore, it is extremely important to obtain the husband's genotype if possible.

Cord blood of all neonates should be assessed, since the presence of SCD in the newborn is associated with a high death rate during early childhood. The early educational value of such a program is immense, since the morbidity and mortality of HbS-S infants seem to be less than when the diagnosis is made during childhood. If cord blood assessment programs are used, then the need for further screening of adults would be eliminated.

Intrauterine Diagnosis.

With the advent of amniotic fluid, chorionic villus and fetal blood sampling, the prenatal diagnosis of fetal hemoglobinopathies offers an extension of genetic counseling. Identification of sickle hemoglobin in the fetus was originally confined to the use of the Hpa I restriction fragment length polymorphism (RFLP) specific for the given genotype, which depended on pedigree studies. Since that time, more specific RFLP and oligonucleotide probes have made in utero diagnosis more accurate, while the use: of polymerase chain reaction with DNA amplification or hemoglobin electrophoresis of fetal hemolysate allow diagnosis within 24 to 48 hours.45,46

Unknown Causes

In some patients, the disorder paroxysmal nocturnal hemoglobinuria (PNH) is diagnosed. This is an anemic disorder associated with marked hypercoagulability of the patient's blood with thrombocytopenia and granulocytopenia but with no anti-platelet antibodies. This disorder is believed to be caused by an abnormal clone of bone marrow stem cells that produces red cells that are extremely sensitive to complement-mediated intravascular hemolysis. Laboratory evaluation reveals mild pancytopenia with an elevated corrected RC. The urine contains hemosiderin, and the appropriate diagnostic assessment is the sucrose lysis test. In these patients, the fertility rate is believed to be depressed and there is also a higher rate of spontaneous abortion. If early abortion does not occur, however, a successful fetal outcome is generally the rule. The major life-threatening events in the mother include postpartum thromboembolism and hepatic vein or cerebral vein thrombosis. The treatment is usually intermittent prophylactic transfusion with washed red cells. The PCV should be maintained between 25% and 30%. In the mother, anticoagulation therapy during the antepartum period and prophylactically during the puerperium seems to be helpful.

Back to Top

1. Bolton FG, Street MJ, Pace AJ: Changes in erythrocyte volume and shape in pregnancy. Obstet Gynecol Surv 38: 461, 1983

2. Goodlin RC: Why treat physiologic anemias? J Reprod Med 27: 639, 1982

3. Kitay DZ: Assessing anemia in the pregnant patient. Contemp Ob/Gyn 2: 17, 1973

4. Fleming AF, Martin JD, Hahnel R: Effects of iron and folic acid antenatal supplements on maternal haematology and fetal well being. Med J Aust 2: 429, 1974

5. Wallerstein R: Iron metabolism and iron deficiency in pregnancy. Clin Haematol 2: 453, 1973

6. de Leeuw NKW, Brunton L: Maternal hematologic changes, iron metabolism, and anemias in pregnancy. In Goodwin JW, Godden JO, Chance GW et al (eds): Perinatal Medicine, pp 425–446. altimore, Williams & Wilkins, 1976

7. Bjorn-Rasmussen E: Iron absorption: Present knowledge and controversies. Lancet 1: 914, 1983

8. Finch CA, Huebers H: Perspectives in iron metabolism. N Engl J Med 306: 1520, 1982

9. Horowitz JJ, Laros RK: Anemia and pregnancy: A review of the pathophysiology, diagnosis, and treatment. J Contin Educ Obstet/Gynecol 29 (2): 9, 1979

10. Morrison JC, Roberts WE: Evaluation of anemia in pregnancy. In Kitay DZ (ed): Hematologic problems in pregnancy. Philadelphia, McGraw Hill, 1985

11. Abramowicz M: Oral iron. Med Lett 20: 45, 1978

12. Buytaert G, Wallenburg HC, van Eijck HG et al: Iron supplementation during pregnancy. Eur J Obstet Gynaecol Reprod Biol 15: 11, 1983

13. Kanakaraddi VP, Hoskatti CG, Nadig VS et al: Comparative therapeutic study of injections of iron dextran complex in anaemia. J Assoc Physicians India 21: 849, 1973

14. Morrison JC: Anemias and hemoglobinopathies. In Ryan GM Jr (ed): Ambulatory Care in Obstetrics and Gynecology, pp 205–238. New York, Grune & Stratton, 1980

15. Morrison JC: Red blood cell disorders. In Gleicher N (ed): Principles of Medical Therapy in Pregnancy. New York, Plenum Press, 1985

16. Kazazian HH Jr: Prenatal diagnosis of β-thalassemia. Semin Perinatol 15: 15, 1991

17. Kitay DZ: Folic acid and reproduction. Clin Obstet Gynecol 22: 809, 1979

18. Johan E, Magnus EM: Plasma and red blood cell folate during normal pregnancy. Acta Obstet Gynecol Scand 60: 247, 1981

19. Fleming AF, Martin JD, Hahnel R et al: Effects of iron and folic acid antenatal supplements on maternal 2 haematology and fetal well being. Med J Aust 2: 429, 1974

20. MRC Vitamin Study Research Group: Prevention of neural tube defects: Results of the Medical Research Council Vitamin Study. Lancet 338:131, 1991

21. Krispel JW, Lynch VA, Viele BD: Aplastic anemia in pregnancy: A case report, review of the literature and reevaluation of management. Obstet Gynecol Surv 31: 523, 1976

22. Douglas SW, Adamson JW: The anemia of chronic disorders: Studies of marrow regulation and iron metabolism. Blood 45: 55, 1975

23. Golbus MS, Laros RK: Hemoglobinopathies and hemolytic anemias. In Schulman JD, Simpson JL (eds}: Genetic Disease in Pregnancy: Maternal Effects and Fetal Outcome, pp 89–122. New York, Academic Press. 1981

24. Dalie JV: Autoimmune hemolytic anemia. Arch Intern Med 135: 1293, 1975

25. Kaplan BH, Bank A: Synthesis of heme: The synthesis of globin. In Williams WJ, Beutler E, Erslen AJ et al (eds): Hematology, 3rd ed, pp 287–300. New York, McGraw Hill, 1983

26. Ferrell H: Hemoglobinopathies: A review of nomenclature. Va Med Month 97: 509, 1970

27. Motulsky AF: Frequency of sickling disorders in U.S. blacks. N Engl J Med 288: 31, 1973

28. Nalbandian RM, Henry RL, Wolf PL et al: Molecular basis for the specific test for hemoglobin S (Murayama test). Ann Clin Lab Sci 1: 26, 1971

29. Perkins RP: Inherited disorders of hemoglobin synthesis and pregnancy. Am J Obstet Gynecol 111:120, 197l

30. Pritchard JA, Scott DE, Whalley PJ et al: The effect of maternal sickle cell hemoglobinopathies and sickle trait on reproductive performances. Am J Obstet Gynecol 117: 662, 1973

31. Rao JN, Sur AM: Iron deficiency in sickle cell disease. Acta Paediatr Scand 69: 337, 1980

32. Klotz IM, Haney DN, King LC: Rational approaches to chemotherapy: Anti-sickling agents. Science 213: 714, 1981

33. Hsia YE: Detection and prevention of important α-thalassemia variants. Semin Perinatol 15: 35, 1991

34. Morrison JC, Schneider JM, Whybrew WD et al: Prophylactic transfusion in pregnant patients with sickle hemoglobinopathies: Benefit versus risk, Obstet Gynecol 46:274, 1980

35. Ricks P: Exchange transfusion in sickle cell anemia and pregnancy. Obstet Gynecol 25: 117, 1965

36. Cunningham FG, Pritchard JA: Prophylactic transfusion of normal red blood cells during pregnancies complicated by sickle cell hemoglobinopathies. Am J Obstet Gynecol 135: 994, 1979

37. Morrison JC, Douvas SG, Martin JN et al: Erythrocytapheresis in pregnant patients with sickle hemoglobinopathies. Am J Obstet Gynecol 149: 912, 1984

38. Cunningham FG, Pritchard JA, Mason R: Pregnancy and sickle cell hemoglobinopathies: Results with and without prophylactic transfusions. Am J Obstet Gynecol 62: 419, 1983

39. Koshy M, Burd L, Wallace D et al: Prophylactic red cell transfusion in pregnant patients with sickle cell disease: A randomized cooperative study. N Engl J Med 319: 1447, 1988

40. Powars DR. Sandhu M, Niland-Weiss J et al: Pregnancy in sickle cell disease. Obstet Gynecol 67: 271, 1986

41. Charache S, Scott J, Neibyl J et al: Management of sickle cell disease in pregnant patients. Obstet Gynecol 55: 407, 1980

42. Milner PF, Jones BR, Dobler J: Outcome of pregnancy in sickle cell anemia hemoglobin C disease. Am J Obstet Gynecol 138: 239, 1980

43. Miller JM, Horger ED, Key TC et al: Management of sickle cell hemoglobinopathies in pregnant patients. Am J Obstet Gynecol 141: 237, 1981

44. Werner E J, Stockman JA: Red cell disturbances in the feto-maternal unit. Semin Perinatol 7: 139, 1983

45. Posey YF, Shah D, Ulm JE et al: Prenatal diagnosis of sickle cell anemia: Hemoglobin electrophoresis versus DNA analysis. Am J Clin Pathol 92: 347, 1989

46. Rowley PT, Loader S, Sutera CJ et al: Prenatal screening for hemoglobinopathies. III. Applicability of the Health Belief Model. Am J Hum Genet 48: 452, 1991

Back to Top