**3.5 Inflammation**

278 From Preconception to Postpartum

As already mentioned, different hemostatic modifications are recognized in normal and in PEc pregnancy; however, the exact pattern of these changes in the fetus is still poorly understood. Some authors have reported a decrease in fibrinogen levels, but there were no differences in tPA, PAI-1 and D-dimer (Zanardo et al., 2005), while others reported an increase in PAI-1, however with no differences in tPA values (Roes et al., 2002). Higgins et al. (2000) suggest that infants are somehow protected, at hemostatic system level, since no differences in D-dimers were observed in newborns whose mothers developed PE (Higgins et al., 2000). In a study performed by our group (Catarino et al., 2008b), we also observed similar values of D-dimers. However, and similarly to what we observed in PEc mothers, significantly increased tPA values were found in the newborns of these women (Fig. 3B). Our results suggest that these changes do not arise in response to activation of coagulation, but as a result of endothelial cell dysfunction. Furthermore, we observed a relationship between placental dysfunction and endothelial dysfunction in fetal circulation in PE, suggested by the positive significant correlation that we identified between the levels of tPA

Fig. 3. Maternal (A) and fetal (B) tPA levels in PE and normal pregnancy. (Adapted from

In PE there is evidence of oxidative stress that results from an increased production of oxidizing agents that is not counteracted by antioxidant activity. In fact, an increase in reactive oxygen species, namely in superoxide anion, and a decrease in superoxide dismutase (SOD) activity is observed in trophoblast cells from PEc women (Wang & Walsh, 2001). Different studies demonstrate the development of oxidative stress in PE (Raijmakers et al., 2004; Roberts & Gammil, 2005); either by decreased concentrations of antioxidants, or indirectly by increased lipid peroxidation in maternal circulation, that results from the action of oxygen metabolites, provided by leukocyte activation and/or increased cellular metabolism. The oxidative stress can also take place at the placenta, as a result of the hypoxia/reoxygenation mechanism (intermittent placental perfusion) observed in placentas

The preterm infants have a lower antioxidant capacity and are therefore more susceptible to oxidative stress triggered at birth, which appears to be associated with complications, including retinopathy and bronchopulmonary dysplasia (Saugstad, 2003). However, there are conflicting results considering oxidative stress in newborns. Some studies describe an

and sVEGFR-1 in umbilical cord blood (Catarino et al., 2009a).

of pregnant women with PE (Hung & Burton, 2006).

Catarino et al., 2008b)

**3.4 Oxidative stress** 

There are various mechanisms by which oxidative stress is linked to the inflammatory process. During the inflammatory response production and release of reactive oxygen metabolites occurs leading to oxidative stress. On the other hand, products of oxidative stress, such as those resulting from lipid peroxidation, are considered pro-inflammatory.

PE may be considered as an exacerbated inflammatory condition compared with physiological pregnancy, which appears to contribute to endothelial dysfunction. Initially, there is a localized inflammatory response within the placenta, while in a second phase predominates a systemic inflammatory response. Numerous studies have reported an increase of pro-inflammatory cytokines such as IL-6 and TNF-α (Bernardi et al., 2008b; Guven et al., 2009; Ouyang et al., 2009) in pregnant women with PE, when compared with normotensive pregnant women.

Concerning acute phase proteins, C-reactive protein (CRP) is probably the most studied one, mainly due to its sensibility in detecting inflammation, with a significant increase being observed in PEc pregnancy (Belo et al., 2003; Tjoa et al., 2003; Guven et al., 2009). Tjoa et al. reported an increase in plasma concentration of CRP between 10 and 14 weeks of gestation in pregnant women who subsequently developed PE and gave birth to newborns with growth restriction (Tjoa et al., 2003).

Cell adhesion molecules (CAM), necessary for the adhesion of leukocytes to vascular endothelium, are also altered in PE. Despite some contradictory results, plasma levels of (sICAM)-1, soluble vascular cell adhesion molecule (sVCAM)-1, soluble platelet endothelial cell adhesion molecule (sPECAM)-1 and soluble E-selectin are raised in PEc pregnant women (Kim et al., 2004; Chavarrνa et al., 2008).

The neutrophil activation also seems to be associated with PE, as some studies mentioned an increase in circulating levels of myeloperoxidase (Mellembakken et al., 2001; Gandley et al., 2008) and elastase (Belo et al., 2003; Gupta et al., 2006), both released during the degranulation of neutrophils in the inflammatory process.

The increase in inflammatory markers in the maternal circulation could result from the release of substances from the placenta (local inflammation), then triggering a systemically inflammatory response. Some authors have suggested that tissue hypoxia resulting from

Umbilical Cord Blood Changes in Neonates from a Preeclamptic Pregnancy 281

In PE there is a change in the lipoprotein subclasses profile, particularly a predominance of small and dense LDL (Sattar et al., 1997). The increase of small, dense LDL fraction is especially important, since this is considered the most atherogenic and also more susceptible to oxidation (Wakatsuki et al., 2000), resulting in the formation of oxidized LDL. The oxidized LDL appears to play a crucial role in endothelial (dys)function observed in PE and some authors report that PE is associated with an increase in oxidized LDL levels (Uzun et al., 2005; Kim et al., 2007). However, other articles referred that there are no significant differences in relation to normal pregnancy (Belo et al., 2005; Qiu et al., 2006). Since oxidized LDL is immunogenic, the formation of autoantibodies to oxidized LDL occurs in circulation; some authors also confirm the increase of autoantibodies to oxidized LDL in the circulation of PEc pregnant women (Uotila et al., 1998). In contrast, other study report no changes (Jain et al., 2004). Additionally, it was proposed that pregnant women who present an increase in oxidized LDL plasma levels are associated with increased risk of developing PE (Qiu et al., 2006). Most changes found in lipid profile, including increased plasma TG and VLDL (TGrich), decreased HDLc and an increase in small, dense LDL subfraction represent a risk profile similar to that predisposing to atherosclerosis and cardiovascular disease (Crowther, 2005). Some authors also propose that changes in lipid metabolism may contribute to the endothelial dysfunction, a key step in the atherosclerotic process (Bayhan et al., 2005; Ray et al., 2006).

Hypertriglyceridemia may contribute to endothelial dysfunction, but may also reflect placental dysfunction, since, unlike free fatty acids and cholesterol, TG does not cross the placenta and have no receptors in the placenta (unlike what happens with lipoproteins). For this, it is necessary that the lipoprotein lipase, which is abundant in the placenta, ensures the TG hydrolysis to be transferred in the form of free fatty acids to the fetus. It is possible that this hydrolysis is impaired in PE, causing a TG accumulation in maternal blood and a reduction in the nutrients uptake by the fetus. On the other hand, the significant increase of TG in maternal blood can be regarded as a physiological mechanism to increase the nutrients supply to the fetus, take into consideration the greater difficulty of nutrients transfer through the placenta. Considering that in PE also occurs a change in placental perfusion, fetal lipid profile may also be affected due to disturbances in placental transfer of lipids. Rodie et al. (2004) reported an increase in TG levels, total cholesterol and total cholesterol/HDLc ratio in newborns of pregnant women who developed PE, although no correlation between the lipid and lipoprotein levels between mothers and newborns was observed. In turn, Ophir et al. (2006) found no differences in TG or total cholesterol, only observed an increase of LDLc in umbilical cord blood from PEc pregnancy. In our previous study, newborns of mothers with PE showed decreased levels of lipids and lipoproteins (exception for HDLc), but a

As already noted, the increase of TG in maternal blood can result from a physiological mechanism of compensation. The high levels of TG in newborns of mothers with PE may,

Some authors argue that to compensate the difficulty of transferring these nutrients will notice an increased expression of a particular type of receptor, for example to LDL. Other authors propose that there is an increased blood flow to overcome the difficulty in transferring nutrients and/or gas and that this adaptation is that justifies the increase in

significant increase in TG (Fig. 4B) (Catarino et al., 2008a).

maternal blood pressure characteristic of PE.

therefore, be a reflection of increased values in the maternal circulation.

reduced placental perfusion, determines an unregulated production of different cytokines, including TNF-α, which is reflected in an increase in the maternal circulation (Rusterholz et al., 2007; Redman & Sargent, 2009).

There is little information for the assessment of inflammatory markers in newborns. Braekke et al. (2005) found no evidence of inflammation in cord blood of newborns from pregnant women with PE, because no differences in CRP or in calprotectin where detected. For studies addressing CAM, the existing information is somewhat contradictory, as is the case for oxidative stress. Some authors have reported an increased expression of L-selectin and integrins on the surface of neutrophils (Mellembakken et al., 2001; Saini et al., 2004) in the fetal circulation in PEc pregnancy, demonstrating an activation of neutrophils. It was also described a decrease in sL-selectin and sE-selectin in the fetal circulation of PEc pregnancy. However, other researchers found no differences between the fetal circulation to a normal pregnancy and PE on levels of sICAM, sVCAM and sE-selectin (Kraus et al., 1998).

The marked inflammatory response in maternal circulation, in the case of PE, seems to be also accompanied by increased inflammatory markers at umbilical cord blood level. Both CRP and α1-antitrypsin are elevated in cord blood, suggesting the presence of an inflammatory response, although less intense than in the maternal circulation (unpublished data).

#### **3.6 Lipid profile**

The physiological hyperlipidemia observed in healthy pregnant women is further exacerbated in PE. PE is characterized by intense changes of lipid profile (Ray et al., 2006) similar to what happens in atherosclerosis. Several studies indicate a significant increase in serum triglycerides (TG) in PEc pregnancy, compared with normal pregnancy (Belo et al., 2002b; Baksu et al., 2005b; Bayhan et al., 2005). This is probably the most consistent finding in lipid profile. In a study performed by our group the most pronounced lipid modification that we found was for TG levels (Fig. 4A), which doubled its value in PE in relation to normal pregnancy (Catarino et al., 2008a). Furthermore, TG levels correlated positively and significantly with proteinuria, a known marker of PE severity (Catarino et al., 2008a). In agreement with this, free fatty acids also appear to be higher in PE (Villa et al., 2009).

It is also referred an increase in LDLc (Bayhan et al., 2005) and total cholesterol (Bayhan et al., 2005) and a decrease in hight HDLc (Belo et al., 2002b; Baksu et al., 2005b; Bayhan et al., 2005) in PE compared with normal pregnancy. However, these parameters are not always altered in PE (Baksu et al., 2005b; Manten et al., 2005) and, when they do, the extent of modification is not so pronounced as with TG.

Another biochemical parameter also subject to some controversy is the lipoprotein (a) [Lp(a)]; some studies reported no significant differences in Lp(a) levels in PEc pregnancy (Belo et al., 2002c; Baksu et al., 2005b), while others described an elevation of Lp(a) in PE (Bar et al., 2002; Bayhan et al., 2005). Moreover, Mori et al. described a positive correlation between maternal Lp(a) levels and the severity of PE (Mori et al., 2003).

reduced placental perfusion, determines an unregulated production of different cytokines, including TNF-α, which is reflected in an increase in the maternal circulation (Rusterholz et

There is little information for the assessment of inflammatory markers in newborns. Braekke et al. (2005) found no evidence of inflammation in cord blood of newborns from pregnant women with PE, because no differences in CRP or in calprotectin where detected. For studies addressing CAM, the existing information is somewhat contradictory, as is the case for oxidative stress. Some authors have reported an increased expression of L-selectin and integrins on the surface of neutrophils (Mellembakken et al., 2001; Saini et al., 2004) in the fetal circulation in PEc pregnancy, demonstrating an activation of neutrophils. It was also described a decrease in sL-selectin and sE-selectin in the fetal circulation of PEc pregnancy. However, other researchers found no differences between the fetal circulation to a normal pregnancy and PE on levels of sICAM, sVCAM

The marked inflammatory response in maternal circulation, in the case of PE, seems to be also accompanied by increased inflammatory markers at umbilical cord blood level. Both CRP and α1-antitrypsin are elevated in cord blood, suggesting the presence of an inflammatory response, although less intense than in the maternal circulation (unpublished

The physiological hyperlipidemia observed in healthy pregnant women is further exacerbated in PE. PE is characterized by intense changes of lipid profile (Ray et al., 2006) similar to what happens in atherosclerosis. Several studies indicate a significant increase in serum triglycerides (TG) in PEc pregnancy, compared with normal pregnancy (Belo et al., 2002b; Baksu et al., 2005b; Bayhan et al., 2005). This is probably the most consistent finding in lipid profile. In a study performed by our group the most pronounced lipid modification that we found was for TG levels (Fig. 4A), which doubled its value in PE in relation to normal pregnancy (Catarino et al., 2008a). Furthermore, TG levels correlated positively and significantly with proteinuria, a known marker of PE severity (Catarino et al., 2008a). In agreement with this, free fatty acids also appear to be higher in PE (Villa et

It is also referred an increase in LDLc (Bayhan et al., 2005) and total cholesterol (Bayhan et al., 2005) and a decrease in hight HDLc (Belo et al., 2002b; Baksu et al., 2005b; Bayhan et al., 2005) in PE compared with normal pregnancy. However, these parameters are not always altered in PE (Baksu et al., 2005b; Manten et al., 2005) and, when they do, the extent of

Another biochemical parameter also subject to some controversy is the lipoprotein (a) [Lp(a)]; some studies reported no significant differences in Lp(a) levels in PEc pregnancy (Belo et al., 2002c; Baksu et al., 2005b), while others described an elevation of Lp(a) in PE (Bar et al., 2002; Bayhan et al., 2005). Moreover, Mori et al. described a positive correlation

between maternal Lp(a) levels and the severity of PE (Mori et al., 2003).

al., 2007; Redman & Sargent, 2009).

and sE-selectin (Kraus et al., 1998).

modification is not so pronounced as with TG.

data).

al., 2009).

**3.6 Lipid profile** 

In PE there is a change in the lipoprotein subclasses profile, particularly a predominance of small and dense LDL (Sattar et al., 1997). The increase of small, dense LDL fraction is especially important, since this is considered the most atherogenic and also more susceptible to oxidation (Wakatsuki et al., 2000), resulting in the formation of oxidized LDL. The oxidized LDL appears to play a crucial role in endothelial (dys)function observed in PE and some authors report that PE is associated with an increase in oxidized LDL levels (Uzun et al., 2005; Kim et al., 2007). However, other articles referred that there are no significant differences in relation to normal pregnancy (Belo et al., 2005; Qiu et al., 2006). Since oxidized LDL is immunogenic, the formation of autoantibodies to oxidized LDL occurs in circulation; some authors also confirm the increase of autoantibodies to oxidized LDL in the circulation of PEc pregnant women (Uotila et al., 1998). In contrast, other study report no changes (Jain et al., 2004). Additionally, it was proposed that pregnant women who present an increase in oxidized LDL plasma levels are associated with increased risk of developing PE (Qiu et al., 2006). Most changes found in lipid profile, including increased plasma TG and VLDL (TGrich), decreased HDLc and an increase in small, dense LDL subfraction represent a risk profile similar to that predisposing to atherosclerosis and cardiovascular disease (Crowther, 2005). Some authors also propose that changes in lipid metabolism may contribute to the endothelial dysfunction, a key step in the atherosclerotic process (Bayhan et al., 2005; Ray et al., 2006).

Hypertriglyceridemia may contribute to endothelial dysfunction, but may also reflect placental dysfunction, since, unlike free fatty acids and cholesterol, TG does not cross the placenta and have no receptors in the placenta (unlike what happens with lipoproteins). For this, it is necessary that the lipoprotein lipase, which is abundant in the placenta, ensures the TG hydrolysis to be transferred in the form of free fatty acids to the fetus. It is possible that this hydrolysis is impaired in PE, causing a TG accumulation in maternal blood and a reduction in the nutrients uptake by the fetus. On the other hand, the significant increase of TG in maternal blood can be regarded as a physiological mechanism to increase the nutrients supply to the fetus, take into consideration the greater difficulty of nutrients transfer through the placenta.

Considering that in PE also occurs a change in placental perfusion, fetal lipid profile may also be affected due to disturbances in placental transfer of lipids. Rodie et al. (2004) reported an increase in TG levels, total cholesterol and total cholesterol/HDLc ratio in newborns of pregnant women who developed PE, although no correlation between the lipid and lipoprotein levels between mothers and newborns was observed. In turn, Ophir et al. (2006) found no differences in TG or total cholesterol, only observed an increase of LDLc in umbilical cord blood from PEc pregnancy. In our previous study, newborns of mothers with PE showed decreased levels of lipids and lipoproteins (exception for HDLc), but a significant increase in TG (Fig. 4B) (Catarino et al., 2008a).

As already noted, the increase of TG in maternal blood can result from a physiological mechanism of compensation. The high levels of TG in newborns of mothers with PE may, therefore, be a reflection of increased values in the maternal circulation.

Some authors argue that to compensate the difficulty of transferring these nutrients will notice an increased expression of a particular type of receptor, for example to LDL. Other authors propose that there is an increased blood flow to overcome the difficulty in transferring nutrients and/or gas and that this adaptation is that justifies the increase in maternal blood pressure characteristic of PE. of

Umbilical Cord Blood Changes in Neonates from a Preeclamptic Pregnancy 283

erythrocyte differentiation, proliferation and the early release of reticulocytes in peripheral blood. The decreased affinity for oxygen of maternal hemoglobin, caused by increased 2,3 diphosphoglycerate within erythrocytes, represents a compensation mechanism for the increased oxygen consumption needed for oxygen transfer to fetus. This transfer is also favoured by the higher oxygen affinity of fetal hemoglobin, which is the predominant form

To respond to the erythropoietic stimulation, there is a mobilization of maternal iron stores and an increase in intestinal iron absorption (Gordon, 2002). To overcome this increase in iron demands during pregnancy, iron supplements are usually given to maintain maternal iron stores, usually after 20 weeks of gestation (Cunningham et al., 2005). During normal pregnancy there is an increase of younger erythrocytes, as shown by an increased number of circulating reticulocytes and, consequently, an increase in "red cell distribution width" (RDW) (Shebat et al., 1998). Tissue hypoxia that occurs physiologically during pregnancy appears to stimulate erythropoietin production, resulting in reticulocyte release from bone

The human erythrocyte has a lifespan of about 120 days, being removed from the circulation, mainly, by the spleen. Mature erythrocyte has no nuclei and organelles, and presents a very limited biosynthetic capacity, accumulating physical and/or chemical changes throughout its life span. Several modifications occur with cell aging, namely,

The erythrocyte membrane protein band 3 is a transmembrane protein, also known as an

transport of CO2 from the tissues to the lungs. Band 3 links the cytoskeleton to the membrane lipid bilayer, participating in the maintenance of cell morphology. Band 3 is also

The development of oxidative stress may occur when exogenous oxygen metabolites diffuse through the membrane or when oxygen metabolites result from autoxidation of hemoglobin, and the red blood cell (RBC) is unable to detoxify the cell, due to depletion in antioxidant defenses. Accumulation of oxygen metabolites can cause hemoglobin oxidation that has a high affinity for the cytoplasmic domain of band 3 (Fig. 5) (Low et al., 1985). This linkage causes band 3 oligomerization and/or aggregation (Waugh et al., 1987), which is recognized by natural autoantibodies anti-band 3 IgG (Fig. 5). The autoantibodies anti-band 3 have a higher affinity for Band 3 oligomers than for band 3 monomers (Lutz, 1992). The band 3 aggregates will act as a neoantigen on the erythrocyte membrane surface, marking the aged or injured erythrocyte for removal by macrophages of the reticulo-endothlial system. Thus, an increase in membranebound hemoglobin (MBH) (Santos-Silva et al., 1998) and in band 3 aggregation are good

Changes in band 3 profile (% of band 3 monomers, aggregates and proteolytic fragments), besides that associated to erythrocyte aging, were also reports in different inflammatory models, namely, in myocardial infarction (Santos-Silva et al., 1995), ischemic stroke (Santos-Silva et al., 2002) and in high competition physical exercise (Santos-Silva et al., 2001). In

3 and Cl-

ions, which is important for the

reduction in cell volume, enzyme activity, antioxidant capacity and deformability.

involved in the removal of senescent/damaged erythrocytes (Wang, 1994).

of fetal hemoglobin.

marrow into the bloodstream (Lurie & Mamet, 2000).

anion channel, as it mediates the exchange of HCO-

markers of erythrocyte senescence and/or damage.

**3.7.2 Erythrocyte membrane protein band 3** 

Fig. 4. Maternal (A) and UCB (B) TG levels in PE and normal pregnancy (N). (Adapted from Catarino et al., 2008a)

The impact of the changes in lipid profile in the future of newborns from PEc mothers is uncertain. However, the increase in TG, the ratio LDLc/HDLc and Apo B/Apo AI, suggest that these infants present an increment in their cardiovascular risk.

#### **3.7 Hematologic system**

When compared with normal pregnancy PE presents an exacerbation of inflammatory and oxidative stress markers. The release of mediators resulting from inflammatory cell activation can trigger changes in surrounding cells. These cell activation products may contribute to erythrocyte damage, accelerating their aging process and its premature removal. This hypothesis can be tested by evaluating markers of erythrocyte production, damage/aging and removal. from

#### **3.7.1 Plasma volume/erythrocyte number**

Throughout pregnancy, plasma volume increases gradually, reaching its maximum at about 30 weeks of gestation. This increase would correspond to about 50% of the average plasma volume in non-pregnant woman (Gordon, 2002) and is essential to face the decrease in vascular resistance within the feto-placental unit, protecting the mother and the fetus from hypotension. It is also important in case of bleeding during the delivery (Gordon, 2002). The erythrocyte number increases progressively until the end of pregnancy. This increase may reach 18% in pregnant women without iron supplementation or 30% when diet is accompanied by iron supplements (Gordon, 2002).

Since plasma volume increases at an earlier stage of pregnancy and more rapidly than the increase in the erythrocyte number, the hematocrit decreases until the end of the second trimester. This hemodilution leads to anemia, commonly known as "physiologic anaemia of pregnancy". Therefore, only when pregnant women present a hematocrit below 0.33 l/l and hemoglobin below 11g/dL, there is a true anemia. From the third trimester, when the increase in plasma volume is accompanied by an equivalent increase in erythrocyte number, the hematocrit stabilizes or increases slightly until the end of pregnancy. During pregnancy, in response to the increased "turnover" of hemoglobin, due to the increased demand of oxygen consumption, there is an increased synthesis of erythropoietin (Gordon, 2002) - a specific erythropoiesis growth factor. Erythropoietin acts at bone marrow, stimulating the

Fig. 4. Maternal (A) and UCB (B) TG levels in PE and normal pregnancy (N). (Adapted from

The impact of the changes in lipid profile in the future of newborns from PEc mothers is uncertain. However, the increase in TG, the ratio LDLc/HDLc and Apo B/Apo AI, suggest

When compared with normal pregnancy PE presents an exacerbation of inflammatory and oxidative stress markers. The release of mediators resulting from inflammatory cell activation can trigger changes in surrounding cells. These cell activation products may contribute to erythrocyte damage, accelerating their aging process and its premature removal. This hypothesis can be tested by evaluating markers of erythrocyte production,

Throughout pregnancy, plasma volume increases gradually, reaching its maximum at about 30 weeks of gestation. This increase would correspond to about 50% of the average plasma volume in non-pregnant woman (Gordon, 2002) and is essential to face the decrease in vascular resistance within the feto-placental unit, protecting the mother and the fetus from hypotension. It is also important in case of bleeding during the delivery (Gordon, 2002). The erythrocyte number increases progressively until the end of pregnancy. This increase may reach 18% in pregnant women without iron supplementation or 30% when diet is

Since plasma volume increases at an earlier stage of pregnancy and more rapidly than the increase in the erythrocyte number, the hematocrit decreases until the end of the second trimester. This hemodilution leads to anemia, commonly known as "physiologic anaemia of pregnancy". Therefore, only when pregnant women present a hematocrit below 0.33 l/l and hemoglobin below 11g/dL, there is a true anemia. From the third trimester, when the increase in plasma volume is accompanied by an equivalent increase in erythrocyte number, the hematocrit stabilizes or increases slightly until the end of pregnancy. During pregnancy, in response to the increased "turnover" of hemoglobin, due to the increased demand of oxygen consumption, there is an increased synthesis of erythropoietin (Gordon, 2002) - a specific erythropoiesis growth factor. Erythropoietin acts at bone marrow, stimulating the

that these infants present an increment in their cardiovascular risk.

Catarino et al., 2008a)

**3.7 Hematologic system** 

damage/aging and removal.

**3.7.1 Plasma volume/erythrocyte number** 

accompanied by iron supplements (Gordon, 2002).

erythrocyte differentiation, proliferation and the early release of reticulocytes in peripheral blood. The decreased affinity for oxygen of maternal hemoglobin, caused by increased 2,3 diphosphoglycerate within erythrocytes, represents a compensation mechanism for the increased oxygen consumption needed for oxygen transfer to fetus. This transfer is also favoured by the higher oxygen affinity of fetal hemoglobin, which is the predominant form of fetal hemoglobin.

To respond to the erythropoietic stimulation, there is a mobilization of maternal iron stores and an increase in intestinal iron absorption (Gordon, 2002). To overcome this increase in iron demands during pregnancy, iron supplements are usually given to maintain maternal iron stores, usually after 20 weeks of gestation (Cunningham et al., 2005). During normal pregnancy there is an increase of younger erythrocytes, as shown by an increased number of circulating reticulocytes and, consequently, an increase in "red cell distribution width" (RDW) (Shebat et al., 1998). Tissue hypoxia that occurs physiologically during pregnancy appears to stimulate erythropoietin production, resulting in reticulocyte release from bone marrow into the bloodstream (Lurie & Mamet, 2000).
