LLEMQRHEBLLELCLL LSL’ECTS OF :LZL‘BU GEL Cb ELLA LLL L: LSLL‘ :GmS LLLE N’LL‘LSLTY L335 ELLE LL. WSLLEGUS H ELLE r balmy ;~ Mix.“ :3? State - ' .J g -." T § ' .v -. P 3, .... =. «my :1. J. ‘ ,. 5‘ ”if? "4'! P J.“ / ”‘Fvupf'm‘ ABSTRACT HEMdRHEOLOGICAL ASPECTS OF REDUCED OSMOLARITY BY William M. Porteous II In this laboratory, vascular flow resistance is being studied in Vivo as a function of blood osmolarity. It is therefore desirable to investigate the effect of reduced osmolarity on viscosity. In this work, a small hemodialyzer was used to obtain canine blood samples with osmolarities between 227 and 320 mOsm/l. Flow measure- ments on the samples were made with a capillary viscometer using a 400 micron diameter tube. This study indicates that, at least in vessels of 400 micron diameter or larger, blcod osmolarity has no significant effect on high shear viscosity, while the yield stress increases with decreased osmolarity. This change in yield stress appeared to be related to the hematocrit. However, since shear stresses in vivo are normally quite high, both viscosity and yield stress changes would appear to be unimportant in flow resistance measurements involving small osmolarity changes. Still open to question is the William M. Porteous II effect of osmolarity on flow through capillaries of the size found in the vascular bed. HEMORHEOLOGICAL ASPECTS OF REDUCED OSMOLARITY BY William M. Porteous II A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemical Engineering 1971 § {*4 D? to Mary Beth ii ACKNOWLEDGEMENTS I would like to express my appreciation to Dr. D. K. Anderson for his guidance and patience and to thank Dr. F. J. Haddy for his suggestions. I also extend my gratitude to the Michigan State University Division of Engineering Research, to Mom and Dad, and to Grandpa for financial support. iii TABLE OF CONTENTS Page DEDICATION. . . . . . . . . . . . . . . . ii ACKNOWLEDGEMENTS . . . . . . . . . . . . . iii LIST OF TABLES . . . . . . . . . . . . . . v LIST OF FIGURES . . . . . . . . . . . . . . vi INTRODUCTION . . . . . . . . . . . . . . . 1 BACKGROUND . . . . . . . . . . . . . . . 3 Rheological Models . . . . . . . Effect of Hematocrit on Apparent Viscosity. . Effects of Osmolarity on Viscosity . . . . GUILD EXPERIMENTAL METHODS AND APPARATUS. . . . . . . . 9 RESULTS. 0 O O Q C O O O O I I O I O O O 12 Osmolarity Changes. . . . . . . . . . 12 Hematocrit . . . . . . . . . . . . 12 Limiting Viscosity. . . . . . . . . . 12 Yield Stress. . . . . . . . . . . . 15 pH 0 C O O O O O O O O O O O O O 17 DISCUSSION 0 O O O O O O O O O O O O O O 21 Limiting Viscosity. . . . . . . . . . 21 Apparent Viscosity. . . . . . . . . . 21 SUMMARY AND CONCLUS IONS O C O O I O O O O C O 2 3 RECOMMENDATIONS O O O O O O O O C O C C O O 2 4 BIBLIOGRAPHY O O O O O O O O O O I O O O C 2 5 APPENDIX A O O O O O O O O O O O O C O C 28 iv LIST OF TABLES Page Experimental Osmolarities and Hematocrits . . 13 Experimental Viscosities and Yield Stresses . 14 Experimental pH's and Temperatures . . . . 18 Means and Sample Standard Deviations for Percentage Changes . . . . . . . . . l9 Correlation Coefficients for Control to Low Osmolarity Changes . . . . . . . 20 Figure l. 2. LIST OF FIGURES Flow—Pressure Apparatus Diagram Yield Stress Versus Hematocrit. vi Page 10 16 INTRODUCTION Though the word hemorheology is less than two decades old, the study which it entails, that of blood deformation and flow, is in its second century. Continual progress is being made in describing the flow of blood through both living bodies and inanimate vessels. Work in this area aids in determining pressure requirements for flow. It helps man to interpret and thus better understand the reasons for many cardiovascular disorders. In addition rheological techniques are used to study blood clotting. Hemorheology, by nature, is interdisciplinary and there- fore both engineers and physiologists become involved. Current research efforts at Michigan State Univer- sity include studying the effect of blood osmolarity on vascular flow resistance. These studies are being con- ducted in vivo by altering the osmolarity of the blood entering the gracilis muscle of the dog and observing the resulting change in the perfusion pressure at constant flow. Normally the osmolarity is assumed to affect the resis- tance by constricting or dilating the vessels. However some of the change may be due to changes in viscosity. Reducing the osmolarity of blood might increase the viscosity by causing water to move from the plasma into the erythrocytes. This in turn results in an increased hematocrit and it has been shown that viscosity is an exponential function of hematocrit. It should be pointed out that most studies on viscosity with respect to hematocrit have been done by varying the number of erythrocytes. The effect is not necessarily the same when hematocrit is altered by chang- ing the size of the erythrocyte. Thus in order to aid in the interpretation of in vivo data on osmolarity effects, this study was under- taken to measure the effects of osmolarity on the rheolo- gical properties of blood. To duplicate experimental con- ditions of the in vivo studies, blood samples were taken immediately after the in vivo experiments had been performed. Though the effects on viscosity are probably dependent on tube diameter, this study was limited to the use of a single tube. Data are presented on the apparent viscosity of blood samples with reduced osmolarity. The corresponding data on the apparent viscosity of control samples are also presented for comparison. BACKGROUND RheologicalLModels Newtonian. For laminar flow of a Newtonian fluid one can utilize a constant viscosity in describing the flow resistance. The flow rate through a circular tube is then given by the Hagen-Poiseuille equation: _ NR4AP (1) Q... 8L‘ “‘9 where Q is the volumetric flow rate, R is the radius, AP is the pressure drOp, L is the length, and up is the viscosity. Because blood is a non-Newtonian fluid, one cannot utilize a constant viscosity. Therefore the apparent vis- cosity is sometimes used to describe the flow resistance. The apparent viscosity for flow through a circular tube is given by: 4 u _ NR AP (2) " 8LQ Casson. Though the viscosity of blood is not constant, the use of a yield stress along with a limiting high sheer viscosity, often will describe flow. Reiner and Scott Blair [10] found that for blood the Casson equation: -dv T15 = Tyl5 + ub15 (-EEE)% (3) gave a satisfactory relation between shear stress and shear rate. Here, T is the shear stress, Ty is the yield stress, and “b is the limiting viscosity at high shear rates. When Equation 3 is integrated for flow through circular tubes, one obtains the volumetric flow rate: 4 _ T ;5 T, T 4 Q = I§;AE_ [} ig ¥X) + §J;10 ‘ %I(?x):] (4) 8 L “b w w w Here Tw is the shear stress at the wall and is given by: RAP Tw = 3—3 (5) Substitution of Equation 4 into Equation 2 gives the apparent viscosity of a Casson fluid flowing in a circular tube: _ i _ 16 T % T _ TY 4 u -ub/[ 74;!) + ($1) (7) (6) a W W W own NIH H The Casson equation, which was originally used to describe the flow of India ink [1], is frequently used to describe blood flow. Though the Casson equation pre- dicts no flow when the shear stress at the wall is less than the yield stress, what actually happens at these low stresses is in question. 5 Effect of Hematocrit on Apparent Viscosity Einstein's Equation. Einstein [3] has shown theoretically that for rigid spheres in a suspending liquid the viscosity of the bulk fluid, is related to the Ubr viscosity of the suspending fluid, up, by: “b = up (1 + 250H) (7) where H is the percent by volume of spheres. Note that neither the size nor the number of spheres are explicitly important. Because erythrocytes are neither rigid nor spheres, better results are obtained for blood when the constant in Equation 7 is other than 250. Exponential Equation. An exponential equation which has also been used to correlate viscosity with hema- tocrit [5] is: ub = upexp (AH) (8) where A is an adjustable parameter. This is an empirical equation that was developed by varying the number of erythrocytes and hence the hematocrit. Cell Size versus Number. With either Equation 7 or 8 viscosity has a large dependence on hematocrit. Unfortunately these equations and their respective parameters have only been used to correlate viscosity with hematocrits altered by changing the number of cells. Therefore the corre- lations are not necessarily valid when hematocrit is altered by changing cell size. However because Equation 7 has both erythrocyte size and number as implicit variables, it should describe viscosity when hematocrit is changed by either procedure. Even though Equation 8 is empirical and based on changing the number of erythrocytes, the reason- ing involved in the derivation of Equation 7, lends support to the use of Equation 8 when cell size is altered. In small capillaries (8 micron diameter) the question of cell size versus number has different implica- tions. Because the erythrocytes are larger than the small capillaries, they must deform to flow through them. Increasing the hematocrit through cell size influences the apparent viscosity by altering the ability of the erythrocytes to deform and flow through the capillaries. Thus the change in apparent viscosity is probably not the same as when hematocrit is changed by increasing the num- ber of cells. Effects of Osmolarity on Viscosity Erythrocyte Size and Deformability. One Of the most important ways in which osmolarity affects blood is by changing the erythrocyte size and shape. The normal shape, that of a biconcave disc with rounded edges, may be altered for various reasons. Lowering the osmolarity tends to make the erythrocytes more spherical. This in turn increases the hematocrit which may increase the apparent viscosity as mentioned in the preceding discussion. In addition Murphy [9] has found that the spheri- cally shaped erythrocytes of hereditary spherocytosis are less deformable and thus increase blood viscosity. A similar effect has been found by Schmid-Schonbein et a1. [13] for erythrocytes with reduced size. Using Millipore filters they found that osmotic crenution increased viscosity. With packed erythrocytes Wells and Schmid- Schonbein [16] also found an increase in viscosity with extreme increases or decreases in osmolarity. Thus large alterations in erythrocyte size and/or shape seem to reduce deformability and hence increase viscosity. Electrokinetic Charge and pH. Blood pH may also be altered when the osmolarity is reduced. According to Rand et a1. [11], pH changes may alter plasma water content enough to affect plasma viscosity, and to a minor but measurable degree, blood viscosity. However no such rela- tion was found by Masin [6]. Both osmolarity and pH affect the electrokinetic charge of cells and decreasing the charge of the erythrocytes was found by Seman and Swank [15] to increase the viscosity. Another factor affecting the electrokinetic charge is that of relative ion concentration. Masin [6] has found that certain ions, in particular potassium, may have an effect on the apparent viscosity. Reduced osmolarity may also affect the charge of proteins. It is not known whether this will in turn affect viscosity. However, Mayer et a1. [7] have studied various proteins, including fibrinogen, and found that protein concentration correlated with viscosity. A possible reason for the electokinetic charge affecting the apparent viscosity is its ability to affect the deformability of and aggregations of erythrocytes. Previous ResultS. Using the Casson relation- ship, Meiselman et a1. [8] studied the effect of increased osmolarity on blood viscosity. They found an increase in limiting viscosity and a decrease in yield stress. A similar study was later performed by Schmid-Schonbein et al. [13] and they confirmed these results. Schmid-Schonbein et a1. [14] also used ultrafiltration to increase osmolarity and reported an increase in apparent viscosity. The effect of reduced osmolarity on blood vis- cosity was also studied by Meiselman et a1. They found an increase in yield stress that was related to the increased hematocrit. The limiting viscosity showed a small decrease. However using a viscometer with shear rates greater than 400 sec.1 Cox and Su. [2] found no change in limiting viscosity with reduced osmolarity. Because Meiselman et a1, made no measurements with shear rates greater than 25 sec-1 these results are not necessarily in conflict. EXPERIMENTAL METHODS AND APPARATUS A hemodialyzer of the type described by Grimsrud and Babb [4] and by Roth [12] was used to decrease the osmolarity of blood samples. The dialyzer was fed blood from the femoral artery of a dog using a constant displace— ment pump. Normal Ringer's solution was used as a dialysate for control samples. Ringer's solution with a reduced NaCl concentration was used to produce low osmolarity blood samples. After the blood had passed through the dialyzer, 50 cc samples were collected in polyethylene bottles which were capped after collection. Within two hours of collection samples were treated with 0.0025 g of heparin to prevent coagulation. One sample was then refrigerated while pH and flow measurements were made on the other at room temperature. The second sample was allowed to come to room temperature and measured immediately after the first. Each run took about 3 hours. The procedure and apparatus (see Figure l) for determining the viscosity were those used by Masin [6] and in short are as follows. A capillary viscometer with a tube diameter of 397.9 microns and a length of 15.796 cm 10 .EmummHQ #19. 023...... msumummmd whammmumusoam .H wusmfim hmoa 923...: mmm A! >¢<4.__n_ u_muzamoz._.< .Illv whisimmizai E mind woz_m>m 20mm .Tull szm> unmm 11 was used. Blood was placed in both upstream and down- stream reservoirs. The upstream reservoir was fed blood using a 1.0 cc syringe and a syringe drive. The pressure drop was measured with a pressure transducer, and voltage output was recorded continuously. Thus for a series of flow rates the pressure drop was measured. The results were then fit by a least squares technique to the Casson equation in integrated form for flow through circular tubes. Thus a limiting viscosity and a yield stress were obtained for eight low osmolarity blood samples and the corresponding controls. RESULTS Osmolarity Changes The average osmolarities for the control and low osmolarity samples were 306 mOsm/l and 265mOsm/l respec- tively (see Table l). The mean percent change in osmolarity was -l3.4 and the standard deviation was 5.5. Hematocrit The percent change in the hematocrit was always positive (see Table l) and had a mean of 6.2 and a standard deviation of 3.8. The correlation coefficient of percent change in hematocrit with percent change in osmolarity was -0.76. Thus a significant change in hema- tocrit and correlation with osmolarity was observed. Limiting Viscosity I The limiting viscosity decreased in five of the eight low osmolarity samples (see Table 2). The mean of the percent change in limiting viscosity was -0.60 and the standard deviation was 15.5. The percent change in limiting Viscosity and the percent change in osmolarity had a correlation coefficient of -0.029. Thus the limiting 12 13 mm.m mm mm ma.b I Ham mam m mo.H m.mv av hm.m I mmm «am 5 mm.m mm mm H.HHI 5mm mmm m vH.h om mm w.oal mom 5mm m mv.m om mm n.man mom mam v em.m mm mm m.m I mum mom m m.~a mm mm m.wm| nmm oom N m.oa mm sv m.hH| mmm mom H mmcmno muHHwHoEmo Honucoo mmcmzo muwumaosmo Houucoo pom usmoumm 30A usmoumm 30A sumo mufluoouwsmm AEmOEV mmfiuflumaosmo .mufluooumsmm paw moauflumaofimo HmucmeflummeII.H mqmms l4 .umuflm consummfi mHmEMm muflHmHOEmo 30H mmuMOHUcH« m.mm mom.o mma.o m.m I v0.5 mm.m m 0.0 moH.o moa.o m.NNI mo.m mv.m n m.mm mmo.o mho.o o.v I m>.v mm.v am n.>¢ onH.o mu .0 h.HN Hm.m ww.m «m m.s I ovm.o mmm.o o.m NH.m mm.m «v 0.0m mmm.o mmm.o H.~ I ow.n om.s m m.va ovo.o mmo.o w.¢HI om.m m¢.¢ m mmm woo.o . mmo.o m.mH o>.m mh.v H moccao mpflnmaosmo Honpcoo mmsmco huHHMHOEwo Houusoo umm usmoumm 30A usmoumm 30A sumo ANE0\mc>©V mommmupm pamfiw Amov mmflunmoomfl> mcfluwssq .mmmmmHflm CHOHW UCM mOHUHmOUmMNV HMHGOEHH0%MII.N mqmdfi 15 viscosity showed no significant change or correlation with osmolarity. Yield Stress The yield stress increased with reduced osmolarity in all but one case (see Table 2). The mean of the percent change was 47.4 while the standard deviation was 78.3. The correlation coefficient with percent change in osmolarity was only -O.214. Thus a significant change in yield stress does not seem to correlate well with osmolarity. However the change in hematocrit and the change in logarithm of the yield stress had a correlation coefficient of 0.74. Thus though the change in yield stress does not seem to correlate directly with the change in osmolarity, there is probably an indirect correlation based on hematocrit. When logarithm of the yield stress versus hemato- crit was fit by a least squares straight line to all data points (both control and low osmolarity) a slope of 0.062 was obtained. The correlation coefficient was 0.65 with a p of 0.04 (see Figure 2). Thus yield stress also had an overall dependence on hematocrit. The average slope of the individual changes in logarithm of the yield stress versus hematocrit was 0.083. A t-distribution was applied to this average and the least squares slope fell within the 75 percent confidence interval. Thus the individual yield stress changes agree with the expected change due to hematocrit. 16 .ufluooumsmm m5mum> mmmuum pamflw .N musmflm uHHOOpmamm om om ow on T _ A _ \ no.0 .. IQ \ I I mncflom mcflpcommmuuoo a a mom mafia maauomccooIlllmv. \ I. o a mo 0 t a i to x \ nu Ex I To “Hm mmumsqm ummmq mHQEmm mpHMmHoEmo 3043 I mHmEmm HOHuGOUG .010 B\ I m.o Zmo/seufip sselqs pTGTA go mquJQSOq 17 £11 With the exception of one case, the pH increased with reduced osmolarity (see Table 3). The increase was significant since the average was 0.134 and the standard deviation was 0.084. However the correlation coefficient with percent change in osmolarity was only -0.186. Thus though the increase in pH appears to be significant, the correlation with osmolarity change is poor. 18 om.H mH.vm mm.mm mo.0I om.s mm.> m «v.0 ms.mm NH.mN mH.o vm.n mH.h s mv.H m.mm wh.om om.o om.h OH.n m mm.H wm.vm wm.m~ «H.o Hm.h hm.> m mm.o mc.Hm H.om Hm.o m¢.s mm.w v so.H mm.mm Nh.Hm HH.o mv.n mm.n m Hm.H mH.mm am.HN no.0 mq.n wm.n m om.H mm.mm mm.om mm.o mm.n mH.h H smcmco Hmch HsHpHcH mmcsco mpHHsHOEmo HOHpcoo usm 30H sumo AUV mononsumosse m.mo .msnsumHsQEsB cos m.mo HsucsEHnmmmeI.m mom