Blood Viscosity of the Neonate

Abstract

After completing this article, readers should be able to: Blood viscosity is defined as resistance to the movement of blood. In a circular vessel or tube, the resistance (R) increases with increasing viscosity (V) of the moving fluid and with the resistance resulting from the vessel geometry (Z): R=Z · V. Thus, blood viscosity describes the contribution of blood rheologic factors to blood flow resistance. However, blood viscosity depends on several factors, and the importance of these factors differs among various vessels. The major determinants of blood viscosity are the hematocrit, plasma viscosity, red blood cell (RBC) aggregation, RBC deformability, leukocyte properties, vessel diameter, and the shear forces acting on the RBC. (1)The hematocrit generally is assumed to be the most important determinant of blood viscosity. Blood viscosity rises exponentially with increasing hematocrit (Fig. 1). A marked rise in blood viscosity is seen when the hematocrit exceeds 0.65 L/L, which typically is defined as the threshold for hyperviscosity of the blood. However, in narrow arteries and arterioles and in capillaries, the plasma viscosity becomes the dominant determinant of blood viscosity. The fetal hematocrit increases from 0.33 L/L in the 12th week of gestation to 0.45 L/L in the 30th week and 0.50 L/L in the 40th week at term. At birth, the hematocrit may rise markedly as a result of “blood transfusion” from the placenta to the neonate. (2) A postnatal hematocrit of 0.45 to 0.65 L/L is considered normal in the healthy term neonate.Plasma viscosity depends on the total plasma protein concentration, but it is influenced more by high-molecular weight proteins (eg, fibrinogen) than by smaller proteins (eg, albumin). The concentrations of total plasma protein, plasma fibrinogen, and other plasma proteins are very low in immature fetuses and preterm infants, increase with gestational and postnatal age, and reach the highest values in adults. (3)(4)(5)RBC aggregation results from bridges formed by macromolecules such as fibrinogen. The aggregation process requires several seconds of contact between adjacent RBCs and, therefore, occurs only during blood stasis or at low shear stresses. At high shear stresses, RBC aggregates are dispersed rapidly. Blood in small preterm infants shows no or very little aggregation during the first minute of stasis. Both the rate and extent of RBC aggregation increase with greater gestational age and are related closely to the fibrinogen concentration. In term neonates, RBC aggregation still is reduced markedly compared with adults. (6)(7)(8)RBC deformability is not a mathematically defined mechanical parameter, but the result of various geometric (RBC shape, volume, and excess surface area) and mechanical (membrane elastic moduli, membrane and internal viscosity) RBC properties. The excess surface area (beyond that required to enclose the cellular volume) determines the maximum extent of RBC deformation. The cell shape determines how much of the excess surface is available for deformation. Neonatal RBCs are markedly larger than adult RBCs, but they have the same excess surface area and swelling capacity. (9) However, blood in preterm and term infants contains more RBCs that have irregular shape (eg, keratocytes, spherocytes, acanthocytes, elliptocytes, echinocytes) than does adult blood. (10)(11) Geometric and mechanical properties of some of these cells deviate markedly from normal discocytes. (11)Elastic moduli define the forces that are required to achieve a given membrane deformation. Three different elastic moduli (bending, shear, and area compressibility modulus) have been determined by means of a micropipette system for neonatal and adult RBC samples. Compared with adults, the elastic moduli of neonatal RBCs were decreased by 16% to 32%. (9)(12) This suggests that the resistance of the neonatal RBC membrane to various types of elastic deformation is lower than that of adults. In preterm infants, the membrane shear elastic modulus is even smaller than in term infants, indicating that their RBC membrane is very flexible. Membrane and hemoglobin viscosity of neonatal and adult RBCs are similar. (12)Despite these favorable membrane mechanical properties, studies on the deformability of neonatal RBCs have conflicting results. The widely used filtration methods showed markedly reduced filterability of neonatal RBCs compared with those of adult cells due to the larger RBC volume in neonates. (13)(14) The pressure required to aspirate a single RBC completely into 3.3-mcm diameter pipettes also was higher for neonatal than for adult RBCs, but the aspiration pressure tended to be lower for neonatal RBCs when cells that had the same volume were compared. (12) Counterrotating devices based on the method of cone-plate viscometry apply well-defined shear forces and are influenced little by the cell volume. Studies using these devices found similar RBC deformability in preterm and term neonates and adults. (15)(16) This may be explained by a balance of favorable (increased membrane elasticity) and unfavorable (more cells that have abnormal shape) properties of neonatal RBCs. A study on RBC deformability in a rheoscope within 1 hour of blood sampling showed significantly increased RBC deformability for preterm and small-for-gestational age infants compared with healthy term neonates who, in turn, showed better RBC deformability than adults. (17)Leukocyte properties are of primary importance for blood flow in small-geometry vessels (ie, capillaries). Leukocytes exert 700 to 1,000 times more resistance to the passage through 5-mcm diameter pores or capillaries than do RBCs. Thus, leukocytes and RBCs provide similar flow resistance in microcirculatory vessels despite the largely different counts. Studies on deformability of neonatal and adult leukocytes showed similar rigidity if the same cell types were compared. (18) Large numbers of rigid immature and activated neutrophils may increase blood viscosity, particularly in narrow vessels. (19)Blood viscosity in neonates increases with rising hematocrit, as in adults. The decrease in plasma viscosity by about 20% in term infants compared with adults and the further 10% to 20% reduction in plasma viscosity in preterm infants results in corresponding decreases in whole blood viscosity in all vessels independent of the vessel diameter and flow velocity. (5)(20)In arteries that have rapid flow, plasma viscosity and RBC deformability are the major determinants of blood viscosity. Deformation of RBCs facilitates bulk flow in relatively wide arteries. In small arteries and arterioles that have diameters of less than 300 mcm, both the hematocrit and the blood viscosity decrease with decreasing diameter (Fåhrjus effect and Fåhrjus-Lindqvist effect, respectively). These effects have been attributed to the migration of RBCs to the vessel center, thereby creating a cell-poor plasma layer on the wall and a cell-rich central core. Inasmuch as the flow velocity increases from the tube wall to the center, the central cell core leaves the vessel more rapidly than the slowly flowing plasma layer at the wall. This results in decreased vessel hematocrit that, in turn, causes a reduction in blood viscosity. Both effects are more pronounced for neonatal RBCs (5)(21) because of their larger volume (mean corpuscular volume [MCV]) and increased membrane and whole cell deformability. (17)The extent of the Fåhrjus-Lindqvist effect increases with rising hematocrit. When flowing from a 500-mcm tube to a 50-mcm tube, viscosity reductions at a hematocrit of 0.70 L/L were 56% in preterm infants, 50% in term neonates, and 39% in adults, whereas the viscosity reductions at a hematocrit of 0.30 L/L were only 35%, 29%, and 19%, respectively. (5) Because of the enhanced Fåhrjus-Lindqvist effect and decreased plasma viscosity, blood viscosity in 50-mcm tubes is similar in neonates who have a hematocrit of 0.70 L/L and adults who have a hematocrit of 0.50 L/L (Fig. 2).Blood viscosity in capillaries that have diameters below the resting RBC diameter depends primarily on the deformability of RBCs. By means of a mathematical model, we calculated that the larger neonatal RBCs require higher pressures for the passage of 3- to 6-mcm diameter capillaries, but that this is compensated for completely by the lower plasma viscosity in the neonate compared with adults. (20)Aggregation of RBCs occurs only at low shear forces in veins and in other vessels that have slow blood flow (eg, after stenoses). Thus, it is likely that the reduced RBC aggregation in neonates (6)(7)(8) facilitates venous return of blood to the right heart. Moreover, blood stasis for a limited time may be less harmful for the neonate than for the adult.In conclusion, the favorable rheologic properties of blood decrease blood viscosity and facilitate blood flow in large and small arteries, arterioles, and veins. These findings may explain why circulation in neonates is less affected by a high hematocrit than is circulation in adults.Table 1 presents hemorheologic and circulatory parameters in preterm and term neonates and adults. Parallel increases of plasma viscosity, systolic blood pressure, and blood flow resistance during maturation and from the term neonate to adult are evident. Vascular hindrance calculated as resistance-to-blood viscosity ratio does not change with greater gestational age, but it increases from the neonatal period to adulthood. Because vascular hindrance describes the effects of vascular geometry on flow resistance, the data indicate that the increase in blood flow resistance with greater gestational age is due to the concomitant increase in blood viscosity. On the other hand, the further increase in flow resistance from the term neonate to the adult can be referred only partially to the moderate increase in blood viscosity (+24%) and is influenced more by the increase in the hindrance (+50%). This simple model suggests that blood viscosity is a major determinant of normal blood flow resistance during maturation of the fetus and during later life.Neonatal polycythemia, defined as a venous hematocrit of at least 0.65 L/L, occurs in about 3% of all newborns. (22)(23) Hyperviscosity, defined as blood viscosity of more than 2 standard deviations above the population mean of screened neonates, occurs in 5% of infants. Table 2 summarizes important causes of and risk factors for polycythemia and hyperviscosity in the neonate. In most neonates, hyperviscosity is a result of polycythemia. The major causes of neonatal polycythemia are increased intrauterine erythropoiesis due to chronic hypoxia and placenta-to-fetus transfusion due to late cord-clamping, keeping the infant below the placenta before cord-clamping, and some placental transfusion due to intrauterine or intrapartum hypoxia. For infants in whom cord clamping is late, the incidence of polycythemia is about 25%. (24)In some disorders, such as maternal diabetes and asphyxia, increased plasma viscosity, decreased RBC deformability, and increased RBC aggregation may contribute to increased blood viscosity in addition to polycythemia. (1) In septicemia, impaired RBC deformability, an increase in rigid immature and activated neutrophils, and increased RBC aggregation may contribute to impaired micro- and macrocirculation. (19)(25)Blood transfuses from the placenta to the neonate when the umbilical cord is clamped at some time after 5 seconds following birth. Before birth, the fetal blood volume is approximately 70 mL/kg. Another 45 mL/kg of blood is contained in the placenta. (2) If the newly born infant is kept at or below the level of the placenta and the umbilical cord is clamped 3 minutes after birth or later, 35 mL/kg of blood may flow into the neonate. The rapid increase in blood volume (50%) is counteracted by extravasation of plasma so that the hematocrit rises from approximately 50% at birth to 65% at 2 to 4 hours after birth. This increase in hematocrit is associated with a rise in blood viscosity by 50% (Table 3). (2) The resulting hyperviscosity may impair blood flow to various organs, thereby compromising their oxygen supply. For this reason, late cord clamping of infants held at or below the level of the placenta has become uncommon. The Leboyer method, which requires placement of the newly born infant on the mother’s abdomen and clamping of the cord when it stops pulsating, is used widely. Thus, the cords of these infants are clamped late, but the pressure gradient between placenta and infant is decreased by lifting the infant above the placenta. Consequently, the volume of placental transfusion and the rise in hematocrit (and blood viscosity) in Leboyer deliveries is in between that seen in infants in whom the cord is clamped early and late. (26) Although plasma proteins are transfused from the placenta to the infant together with RBCs, late cord-clamping was not associated with rising plasma proteins and plasma viscosity. (26) This may be explained by rapid extravasation of whole plasma due to the increased leakiness of capillaries in neonates. (2)Intrauterine hypoxia or enhanced uterine contractions (eg, oxytocin treatment) may cause marked placental transfusion to the fetus and polycythemia before birth. (2)Prolonged intrauterine hypoxemia stimulates erythropoiesis and causes a shift of blood from the placenta to the fetus, thereby increasing the hematocrit. (2) Impaired oxygen supply to the placenta (eg, maternal toxemia, smoking, placental insufficiency), therefore, often is associated with polycythemia of the fetus and the neonate. (27)(28) Moreover, prolonged intrauterine stress may cause a rise in plasma fibrinogen, thereby increasing plasma viscosity and RBC aggregation, and a rise in total leukocyte count and in the percentage of immature and rigid granulocytes. RBC deformability is not altered in infants who experience intrauterine growth retardation compared with healthy neonates. (17)(29) It is interesting to note that hemodilution of mothers who have severe toxemia and a marked rise in hematocrit not only improves the condition of the mother, but also improves the growth of the infant and decreases the hematocrit and blood viscosity to the normal range. (30) Buchan (31) proposes that increased blood viscosity in infants who have intrauterine growth retardation may be an important risk factor for hypertension in adult life.Acute hypoxemia and acidosis may increase the influx of water and ions into the RBC, thereby increasing the MCV and decreasing the excess surface area. RBC membrane properties may be impaired by hypoxia and acidosis, thereby decreasing RBC membrane deformability. Acidosis and hypoxia cause a greater decrease of RBC filtration rate in the fetus and neonate than in adults. (32) Even the moderate hypoxic stress of normal vaginal delivery decreases RBC filterability below that of neonates born by primary cesarean section. Plasma viscosity does not rise as a result of acute fetal hypoxemia. (33)Poorly controlled insulin-dependent diabetes during pregnancy may cause a variety of metabolic problems in the infant due to hyperglycemia and hyperinsulinism. The hematocrit may rise as a result of hyperinsulinism and placental insufficiency. Moreover, plasma viscosity of affected infants may be increased, and RBC deformability may be decreased. (34)(35) The decrease in RBC deformability is probably a result of decreased membrane elasticity (ie, increased membrane elastic shear modulus). (34) These impaired hemorheologic properties may contribute to the increased risk of infants of diabetic mothers for thromboembolic complications.Many preterm infants are treated with erythropoietin to prevent or treat anemia. Uncontrolled treatment may result in polycythemia in some preterm infants who respond particularly well to erythropoietin. Erythropoietin treatment of children who have uremia increases their hemoglobin concentrations and improves the cellular and membrane deformability of their RBCs due to the increased formation of young, well-deformable RBCs. (36) Studies of hemorheologic parameters during erythropoietin treatment of neonates apparently have not been published.Plasma viscosity increases linearly with rising total plasma protein, but it is influenced more by high-molecular weight proteins (fibrinogen, immunoglobulins) than by low-molecular weight proteins. Moreover, macroglobulins increase RBC aggregation due to formation of bridges between adjacent bridges. Many newborns are treated with high doses of immunoglobulin for prophylaxis or treatment of septicemia, Rhesus immune hemolysis, and alloimmune thrombocytopenia. In adults, high doses of immunoglobulins may result in a marked rise of plasma viscosity and RBC aggregation in vitro and in vivo. (37) Thromboses and necrotizing enterocolitis have been observed in neonates who have received high doses of immunoglobulins. (38) Because of the higher plasma viscosity of adult plasma, fresh frozen plasma should not be used for hemodilution in neonates who have polycythemia. (3)Erythrocytes from adult donors are less deformable than are neonatal RBCs. (17) Accordingly, transfusion of adult RBCs should not result in a hematocrit of more than 0.55 L/L. (5)The hematocrit-dependent rise in blood viscosity accounts for clinical manifestations of polycythemia. Increased blood viscosity can increase the flow resistance in various organs, thereby impeding their blood and oxygen supply. Polycythemia in the neonate decreases cardiac output and blood flow to the brain, gastrointestinal tract, kidneys, lungs, limbs, and skin. (39) There is concern that this may increase the risk of pulmonary hypertension, renal failure, necrotizing enterocolitis, cerebral ischemia, intracranial hemorrhage, and developmental retardation. (22)(23) However, systemic RBC transport (calculated as the product of cardiac output times hematocrit) and RBC transport to the brain (blood flow velocity times hematocrit) remain stable in the neonate over a hematocrit range of 0.40 to 0.70 L/L. (40) At a hematocrit of 0.70 L/L or greater, systemic and cerebral RBC transport decrease markedly. Moreover, cerebral oxygenation does not decrease in the neonate up to a hematocrit of 0.70 L/L. (41) These in vivo findings agree with in vitro studies of blood flow in narrow tubes. Transport of neonatal RBCs in tubes that have a diameter of 50 mcm is stable over a hematocrit range of 0.40 to 0.70 L/L, whereas transport of adult RBCs tends to decrease at hematocrits greater than 0.55 L/L. (5)Clinical consequences of a high blood viscosity are principally a result of impaired circulation in affected organs (Table 4). Reported frequencies of clinical manifestations in polycythemic neonates vary widely. This may be explained by: 1) inclusion of neonates who have a high risk of perinatal complications independent of the occurrence of polycythemia (eg, asphyxia, intrauterine growth retardation, diabetic mother, malformations) and 2) different numbers of neonates who have venous hematocrits of 0.70 L/L or greater. Wiswell and associates (42) observed clinical signs and symptoms in 50% of neonates who had venous hematocrits of 0.65 L/L or greater; van der Elst and colleagues (43) found most polycythemic infants healthy and unaffected by hematocrits of at least 0.65 L/L. Moreover, most clinical signs and symptoms of polycythemia are of minor importance and are observed in neonates who do not have polycythemia.Long-term studies of the incidence of developmental and neurologic abnormalities at 1 to 7 years also reach conflicting conclusions (Table 5). Goldberg and associates (44) reported a markedly increased incidence of neurologic problems at 9 months among infants who had polycythemia. Black and colleagues (45) studied infants who had polycythemia with or without hemodilution at 1, 2, and 7 years of age. Compared with a control group that had normal neonatal hematocrits, neonatal polycythemia was associated with a markedly increased risk of neurologic and developmental long-term problems for 2 years. At 7 years of age, only small (insignificant) differences were observed among the three groups. However, both investigations included infants who had additional risks for developmental problems. Moreover, hemodilution did not influence the results significantly. Investigators who included only polycythemic infants without additional risks found no effect of polycythemia and hemodilution on long-term outcome. (43)(47)(48) Drew and associates (49) reported that increased blood viscosity was a better predictor of poor outcome than was a high hematocrit.Neonates who are affected by intrauterine growth retardation, maternal diabetes, asphyxia, or late cord-clamping have a particularly high risk for polycythemia. The most frequent clinical signs of polycythemia are peripheral or systemic cyanosis and plethora. Neonates who are at increased risk or exhibit clinical signs should undergo postnatal blood sampling for hematocrit determination. The hematocrit is markedly higher in skin prick (“capillary”) blood than in venous or arterial blood, particularly if the peripheral blood flow is low. The microcentrifuge hematocrit may be slightly higher (due to trapped plasma of about 2%) than that calculated from RBC volume and RBC count determined by hematology analyser. (50)In most term and preterm infants of 32 to 36 weeks’ gestation, the hematocrit increases after birth, reaches the highest values at 2 hours of postnatal age, and decreases slowly over the next 24 hours. (51)(52)(53) After cord clamping within 5 seconds of birth, the hematocrit does not change during the first 2 hours, but decreases significantly during the following 24 hours. After late cord clamping, the maximum hematocrit also is reached at 2 hours. (24) Thus, the frequency of polycythemia depends strongly on the time of blood sampling.Screening programs for the detection of polycythemia are based primarily on hematocrit measurements in cord blood. An umbilical hematocrit of greater than 0.55 L/L is associated with a high risk of polycythemia at 2 hours, defined as a hematocrit of at least 0.65 L/L. (51)(52)(53) However, healthy term infants who had early and late cord clamping showed similar umbilical hematocrit values, although 25% of infants who had late cord clamping had hematocrits of 0.65 L/L or greater at 2 hours. (24)Because increased blood viscosity is a better predictor of impaired long-term outcome than is increased hematocrit, it makes sense to use blood viscosity measurements to indicate the need for exchange transfusion for hemodilution. Blood viscosity techniques generally are not available. However, blood viscosity in narrow tubes that have diameters of 50 or 100 mcm can be calculated from published formulas. (5)Prevention of hyperviscosity includes prevention of risk factors such as poor control of maternal diabetes and intrauterine asphyxia. If the risk for polycythemia is increased, the umbilical cord should be clamped immediately after birth to avoid any placental transfusion. (54) Polycythemia can be treated only by hemodilution. Isovolemic hemodilution in neonates usually is performed via an umbilical venous catheter using 5% human serum albumin, serum (free of activated clotting factors), or crystalloids (normal saline or Ringer solution). Adult plasma increases the plasma viscosity and the RBC aggregation and, therefore, should not be used for hemodilution. (3) Although crystalloids leave the circulation rapidly, they have been shown to be as effective as protein colloids for reduction of the hematocrit and amelioration of short-term outcome. (55) This may be explained by even dilution of intra- and extravasal proteins by crystalloids. Plasma expanders such as hydroxyethyl starch and Hemaccel have been used for hemodilution in neonates, (56) but little is known about their distribution and metabolism in the neonate. Because the central hematocrit is 0.65 L/L or greater in about 3% of all neonates, several thousand otherwise healthy neonates may be exposed to the potentially harmful invasive procedure of partial exchange transfusion if this threshold is used as an indication, although long-term studies of children who have neonatal polycythemia failed to show the benefits of exchange transfusion for hemodilution. However, the long-term studies summarized in Table 5 include only a few children who had neonatal hematocrits of at least 0.70 L/L. Hemodilution, therefore, is recommended presently if the hematocrit is 0.70 L/L or higher.Our policy is as follows: See also Philip AGS, Saigal S. When should we clamp the umbilical cord? NeoReviews. 2004;5:e142 –e154