ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 22, No. 4 Copyright 1992, Institute for Clinical Science, Inc. Whole Blood Viscosity in Beta Thalassemia Minor JAMES P. CROWLEY, M.D.t, JACLYN B. M ETZGER, B.S.t, EDWARD W. MERRILL, Ph.D., and C. ROBERT VALERI, M.D. 1 tdepartment of Medicine, Rhode Island Hospital, Providence, RI 02903 and 1Boston University School of Medicine, The Naval Blood Research Laboratory, Boston, MA 02218 and Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139 ABSTRACT Patients w ith heterozygous (3-thalassemia m inor have a decreased hem atocrit (HCT). Since the H CT is a primary determ inant of whole blood viscosity, the known reduction in HCT in (5-thalassemia m inor should lead to a m easurable reduction of whole blood viscosity. The influence of the relatively low er m ean corpuscular volume and consequent higher red blood cell count and (3-thalassemia minor on whole blood viscosity using a m icroporous viscom eter has not previously b een the subject of investigation. Accordingly, the blood of a group of normal and (3-thalassemia m inor subjects was examined with a microporous viscometer to elucidate further the relations betw een whole blood viscosity, HCT, and red blood cell count. The data show that for normal and (3-thalassemia m inor subjects a significant positive correlation (r = 0.65, p < 0.01) exists betw een HCT and whole blood viscosity. However, the slope of the regression of whole blood viscosity and HCT of (3-thalassemia m inor subjects was significantly higher z = 3.14, p < 0.001) than that of normals. Thus, for any given HCT their w hole blood viscosity was higher than that of normals. Studies of the rela- * This work was supported by the U.S. Navy Office of Naval Research Contract N00014-88-C0118, with funds provided by the Naval Medical Research and Development Command. * The opinions or assertions contained herein are those of the authors and are not to be construed as official or reflecting the views of the Navy Department or the Naval Service at large. Dr. Crowley is the recipient of a Transfusion Medicine Academic Award from the National Heart, Lung and Blood Institute (K07 HL01259-01A1). t Send reprint requests to: James P. Crowley, M.D., Rhode Island Hospital, 593 Eddy Street, Providence, RI 02903. 229 0091-7370/92/0700-0229 $01.20 Institute for Clinical Science, Inc.
230 CROWLEY, METZGER, MERRILL, AND VALERI tion of red blood cell counts to whole blood viscosity indicate the higher whole blood viscosity at a given HCT was related to the increased red blood cell counts in (3-thalassemia m inor subjects. Because of the opposing interactions of HCT and red blood cell counts, the m ean whole blood viscosity of the group of (3-thalassemia minor subjects examined was not significantly lower than the normal whole blood viscosity. In fact, (3-thalassemia m inor subjects with H CT > 42, had a whole blood viscosity significantly higher (p < 0.001) than normals with HCT > 42. Introduction A recent study of hospitalized patients has dem onstrated a lower than expected in c id e n c e of ^ -th a la sse m ia tra it in patients adm itted to the hospital with m yocardial infarction.2 Since m ild anem ia is often p resen t in subjects w ith (3-thalassemia m inor,7 one explanation for this finding is an ameliorative effect of a lower hematocrit. A beneficial effect of m ild an em ia on th e a th e ro scle ro tic microcirculation m ight be related to lowered whole blood viscosity since a lower w hole blood viscosity w ould lead to im proved blood flow in vessels w here resistance to flow has increased.1 The relation betw een HCT and whole blood viscosity has b e e n studied previously.4 The HCT may be used to estimate w hole blood viscosity because the relative viscosity of blood to w ater varies approximately as the HCT is raised to the power of 1.3. However, since the viscosity of blood increases as the velocity of flow decreases owing to the aggregation of red blood cells, G uyton5 suggested that clinically relevant m easurem ents of w hole blood viscosity require an instrum ent with m easurem ental characteristics resem bling the low shear rates and shear stress of the m icrocirculation. A porous sintered polyolefin viscom e ter has been developed with a low shear rate and shear stress for clinical m easurem ent of whole blood viscosity.3 Using this instrument, the whole blood viscosity was m easu red of 31 adults including 16 subjects w ith (3-thalassemia minor to determ ine if the whole blood viscosity in (3-thalassemia m inor subjects is actually reduced. Materials and M ethods B l o o d Sa m p l e s Fifteen to 20 ml of blood were drawn into a syringe by venipuncture from each in d iv id u a l stu d ie d. S ev en m l w ere placed in a tube containing ethylenedia m in e te tra a c e tic acid (ED TA )* and reserved for hematologic m easurem ents. The rem aining blood was im m ediately used to m easure viscosity. An additional 30 ml of blood was drawn on some individuals and m ixed w ith 4.3 ml anticoagulant citrate phosphate dextrose solutiont to be used for the preparation of saline-suspended red blood cells. Su b j e c t s Thirty-one individuals w ere studied. Fifteen of these subjects w ere normal, healthy hospital personnel w ith no history of anemia and normal HCT values. Only microhematocrits were performed. The rem aining 16 individuals were all previously diagnosed by hem oglobin electrophoresis as having thalassem ia trait. T hirteen of the 16 had elevated hem oglobin A2 levels (8.3 ± 3.1) availa b le th ro u g h h o sp ita l re c o rd s. T he * Vacutainer, Becton Dickinson and Company, t Fenwal Laboratories.
WHOLE BLOOD VISCOSITY IN BETA THALASSEMIA MINOR 231 rem aining th re e had b een diagnosed elsew here years ago and the actual A2 level was unavailable. However, each of these three had a positive family history for (3-thalassemia minor, microcytosis and anem ia. T he first eight (3-thalassemia m inor su b jects w ere te s te d only by m icrohem atocrit. T he last eight in d i viduals had C oulter counts performed, including red blood cell counts and mean corpuscular volum e. V is c o s it y D e t e r m in a t io n 3 Viscosity m easurem ents w ere m ade with a microporous viscom etert which is a disposable device composed of a cylindrical plug of sin te re d polyolefin connected to a capillary tube above and a fitting for connection to a three way Luer lock stopcock below. The m icroporous polyolefin plug is characterized by a Darcy perm eability (B0) of about 8 Darcy (8 x 10-8 cm2) as determ ined by calibration w ith three standard silicone calibration oils ranging from five to 20 centipoise and the use of the defining equation: Q/A = ( B > ) (AP/L) where Q = cm2 per sec of fluid flow tow ard plug; A = total cross sectional area of plug (solid included) normal to approaching flow, cm2 (fixed); (x = fluid viscosity, in poise (dyn/sec/cm-2) (variable); AP = pressure drop due to hydrostatic head (fixed), dyn/cm -2 = pg AHavg. Thus, with AHavg-13.3 cm (the mean hydrostatic head in the capillary tube), g t Merrill Corporation, Cambridge, MA. = 980 cm /sec2, and density p = 1.04 g/cm -3 (blood, blood plasm a, silicone calibration oil), AP is found to be 2.8 x 103 dyn/cm -2. The flow time, in seconds, was found to be approximately 4.0 times the v isco sity in c e n tip o ise, a n d, as expected, was linearly proportional to the viscosity of the calibration liquid. Variation in flow times betw een viscometers with the same calibration fluid was less than ±5 percent. These conditions lead to the result that the effective shear stress in the pores, around one dyn/cm -2 or less, will allow blood to flow slowly enough to reveal its non-newtonian properties, since red cell interactions m ediated by fibrinogen will be significant. The mean diameters of the communicating channels of the porous bed, are within a range relevant to arterioles and venules («40 jjun). M oreover, the low pressure gradient across the bed, in combination with its perm eability and the slow flow rate sim ulates flow in the microcirculatory vessels. The m ean shear stress, about one dyn per cm2, is fixed by the m eniscus height, the b ed cross section, and the pore diameter. The flow rate adjusts itself according to the tendency of the red cells to interact. To m easure viscosity at room tem perature, three microporous viscometers are m ounted vertically, and approxim ately 10 ml of blood are draw n by venipuncture into a syringe w ithout anticoagulation. The needle is rem oved from the syringe, and any air b u b b les in th e syringe are expelled. The syringe is connected to the three-way stopcock at the base of the first microporous viscometer, and the blood is gently injected into the viscometer, flowing upwards through the porous bed and into the capillary tube. W hen the meniscus of the blood is above the upper mark on the tube, the syringe is removed, and the stopcock arm is raised to 90 perm itting the blood to flow dow n wards by gravity through the viscom eter.
232 CROWLEY, METZGER, MERRILL, AND VALERI A stopcock watch is started when the m eniscus reaches the upper mark, and the tim e recorded w hen the m eniscus reaches the lower mark on the capillary tube. The m easurem ent is done in triplicate using a new microporous viscometer for each m easurem ent, and the m ean flow tim e is recorded as the estim ate of viscosity. H e m a t o l o g ic P a r a m e t e r s D ecreased hem atocrit was m easured either by a standard high speed microhem atocrit m ethod or by a Coulter Counter, M odel S.* W hole blood red blood cell counts and m ean corpuscular volum e w e re a lso m e a s u re d b y a C o u lte r Counter, M odel S.* Saline suspensions or red blood cell counts were counted by a standard m ethod on a Coulter Counter, Model ZBÏ.* P r e p a r a t io n o f Sa l in e S u s p e n d e d R e d B l o o d C e l l C o u n t s C itrate phosphate dextrose solution, anticoagulated blood was washed three tim es by centrifugation in a saline solution containing 0.9 percen t NaCl, 0.2 percent dextrose, 40 mg percent inorganic phosphorus, ph 7.0 with an osmolality of 340 mosm per kg. After each centrifugation, care was taken to remove all buffy coat visible. Sufficient saline solution was added to produce a suspension with either a HCT of 0.60 or a 6.0 x 1012 per m l red b lood cell counts. F or each suspension, both m icrohem atocrit and red blood cell counts w ere performed. Viscosity was m easured in duplicate on each sample. St a t is t ic a l A n a l y s is The m ean and standard deviation (SD) or standard error (SE) of the m ean was * Coulter Electronics, Hialeah, Florida. reported for each group in each type of experiment. The means of the two groups w ere com pared using the non-paired t-test. L inear regression, correlation coefficient, analysis of covariance, and the paired t-test w ere used to m easure the relationship b e tw e en w hole blood viscosity and a variable (e.g., HCT). Slopes were compared by dividing the arithm e tic difference betw een the slopes by the confidence limits. A z value of >2.0 was considered significant. C ertain m easurem ents w ere perform ed only on some individuals. T he data indicate the num ber (N) of samples used for each determ i nation. A p value of <0.05 was considered significant for all tests. R esu lts T here was no significant difference betw een the w hole blood viscosity of normal subjects w hen compared to that of p-thalassemia m inor subjects by the independent t-test (Normal = 34.1 ± 6.3 sec., vs. p-thalassemia m inor = 40.0 ± 14.3 sec.). There was a significant correlation (r = 0.65, p < 0.01) betw een HCT and whole blood viscosity overall (figure 1). However, it was apparent that most of the p-thalassemia m inor points lie to the left of the calculated regression line, and most of the normal points lie to the right. There was a significant difference (z = 3.14, p < 0.001) betw een the slopes of the regression lines that can be calculated for each of the groups which indicates that whole blood viscosity rises more as a function of HCT in 3-thalassemia minor subjects than normals. Thus, if the (3-thalassemia m inor and normal subjects are divided into two groups, those with a HCT < 42 and those with a HCT > 42, the m ean of p-thalassemia minor subjects with HCT > 42 is significantly higher (p < 0.001) than that of normals (figure 2). A strong correlation (r = 0.9, p < 0.001) was seen betw een whole blood viscosity and red blood cell counts in p-thalasse-
WHOLE BLOOD VISCOSITY IN BETA THALASSEMIA MINOR 233 (3-thalassemia m inor suspensions had a significantly lower viscosity than did the normal suspensions (figure 4). Discussion PCV (Volumes %) FIG URE 1. T h e r e la tio n b e tw e e n h e m a to c r it (p a c k e d c e ll v o lu m e ) a n d w h o le b lo o d v isc o sity m e a s u re d b y th e m ic ro p o ro u s v isc o m e te r fo r 15 n o r m a l a n d 16 (3 -thalassem ia m in o r su b je c ts. mia m inor subjects (figure 3). The possibility that the red blood cells may have accounted for the higher w hole blood viscosity seen in 0-thalassem ia m inor subjects relative to normal at any HCT was considered. The influence of red blood cells on w h o le b lo o d v isco sity was assessed by washing normal and (3-thalassemia m inor red blood cells, suspending them in saline at a HCT of 60 (see M ethods) and m easuring viscosity. A strong correlation (r = 0.9, p < 0.001) was dem onstrated for viscosity and red blood cells at a fixed HCT (figure 3). Moreover, suspensions prepared from the blood of (3-thalassemia m inor subjects (figure 4) had a higher m ean viscosity (77 ± 8.1 sec.) than did those of normals (56.5 ± 4.8 sec.) w hen HCT was adjusted to a similar value (60 vol percent) for both groups. Conversely, w hen three red blood cell suspensions were prepared and adjusted to the same red blood cell count (6.0 x 109 per ml) and viscosity m easured, the Under conditions of low shear rate and shear stress, the whole blood viscosity of a group of (3-thalassemia m inor subjects was not significantly different from that of normal subjects ((3-thalassemia m inor = 40.0 ± 14.3 sec., Normal = 34.1 ± 6.3 sec.), and at H C T > 42 it was significantly higher. The whole blood viscosity of subjects with (3-thalassemia m inor has be e n studied previously by other tech n iq u es. Stasiw e t a l8 u sed a W ells- Brookfield cone plate viscom eter and also found no major difference in whole blood viscosity b etw een five (3-thalassem ia m inor subjects and six norm al subjects. O ur results are consistent w ith these reports since there is no significant difference betw een w hole blood viscosity of norm al and (3-thalassemia m inor subjects. _ * N o rm al -p /3-thal * p<0.001 n = 9 n = 5 n = 7 n = 1 0 PCV < 4 2 PCV > 42 F ig u r e 2. A c o m p a riso n o f w h o le b lo o d v isc o sity o f 3 -th a la ssem ia m in o r a n d n o rm a l su b je c ts a fte r se p a ra tin g e a c h in to tw o g ro u p s b a s e d o n th e h e m a to c rit (p a c k e d c e ll v o lu m e).
234 CROWLEY, METZGER, MERRILL, AND VALERI RBC x 10 /ml F ig u r e 3. T h e re la tio n b e tw e e n v isc o sity a n d r e d b lo o d c e ll c o u n ts fo r six n o rm a l a n d 11 3-thala s se m ia m in o r su b je c ts a fte r th e re d b lo o d c e lls w e re w a s h e d a n d r e s u s p e n d e d in b u ffe re d s a lin e at a h e m a to c rit o f 60. Tillman and Schroter10 m easured the viscosity of erythrocyte suspensions p repared from the blood of 12 p-thalassemia m inor subjects at a fixed HCT of 80 with a W ells-brookfield viscom eter at shear rates betw een 0.39 and 78.6 sec.-1 and found the values to not be significantly different from normal. In contrast, a significantly higher whole blood viscosity was found by us (figure 3) w hen (3-thalassem ia m inor red blood cells w ere suspended in saline at a HCT of 60. As discussed in detail in our recent report on the microporous viscom eter,3 there are sig n ifican t m ethodologic differences betw een the microporous viscom eter and cone plate viscometers. It is likely that the branching capillary bed of the microporous viscom eter m akes it m ore sensitive to the higher red blood cell counts that result from the low m ean corpuscular volume in p-thalassemia m inor than the flat surface of the cone plate viscom eter. Strumia studied the effect of red cell factors on the relative viscosity of whole blood using an Ostwald U-tube viscometer. He found that the relative viscosity of w hole blood m easured this way was influenced by the red cell size: w hen the red blood cell count was held constant, m icrocytosis was acco m p an ied by a d ecrease of the relativ e viscosity of w hole blood w hile m acrocytosis was accompanied by an increase.9 The difference betw een our results and th at of Strum ia is probably explained by the experimental m ethod, since the Ostwald viscom eter has a very high shear rate relative to the viscom eter used by us. Why the size of the red cell would have the opposite effect on blood studied in a high shear rate versus a low shear rate is unknow n and is deserving of further study. As discussed, our results are consistent with the more recent reports of the w hole blood viscosity in p-thalassem ia m inor8,10 since no significant difference betw een whole blood viscosity of normal and p-thalassemia m inor subjects was observed by us. RBC count of 6.0 x 10s/ml F ig u r e 4. A c o m p a riso n o f th e v isc o sity o f th e w a s h e d, s a lin e s u s p e n d e d re d b lo o d c e lls o f 0-th ala sse m ia m in o r a n d n o rm al su b je c ts.
WHOLE BLOOD VISCOSITY IN BETA THALASSEMIA MINOR 235 W hile the m ean whole blood viscosity of (3-thalassemia m inor subjects was no different from norm al, the correlation betw een H CT and whole blood viscosity was significantly different from that of normal (figure 1, Results). For any given HCT, there was a higher whole blood viscosity in (3-thalassemia m inor than that predicted from normals. The lower mean corpuscular volum e in (3-thalassem ia m inor causes a higher red blood cell count than normal at any given HCT, which suggests that the red blood cell count itself may independently influence the w hole blood viscosity in (3-thalassemia minor. W hen normal and (3-thalassem ia m inor erythrocytes w ere each suspended to a fixed H CT of 60, a highly significant difference in viscosity was dem o n strated b e tw e en (3-thalassem ia m inor and normal (p < 0.001), thus confirming the independent influence of red blood cell counts on whole blood viscosity. The increase in whole blood viscosity owing to the higher red blood cell counts at any given HCT explains the failure to observe the expected reduction in whole blood viscosity for (3-thalassemia minor. A note on technique seems worth m entioning. T he m easu rem e n t of w hole blood viscosity w ith the m icroporous viscometer was a relatively sim ple device to m easure w hole blood viscosity w hen com pared to other existing m ethods.3 Moreover, the viscom eter was sensitive to alterations other than HCT (viz red blood cell counts) that influence whole blood viscosity. Thus, the m easurem ent of whole blood viscosity with the microporous viscom eter deserves further study in settings in which dysproteinem ias or extreme leukocytosis alter the usual and expected relation betw een the HCT and the w hole blood viscosity. The results of this study suggest that a decrease in w hole blood viscosity does not seem a likely explanation for the prior observed reduction in the occurrence of m yocardial infarction in subjects w ith (3-thalassemia m inor.2 An alternative suggestion that the reduced incidence of m yocardial infarction in (3-thalassemia m inor may be due to the low er cholesterol levels needs further investigation.2,3 Studies of high and low density beta lip o p ro te in s or fra c tio n a tio n o f th e apolipoproteins w ould therefore be of special interest in (3-thalassemia m inor in view of the predictive value of these m e a su rem e n ts for th e d e v e lo p m e n t of atherosclerosis.6 References 1. Ba r g e r, A. C., et al: Hypothesis: Vasa vasorum and neovascularization of human coronary arteries. New Engl. J. Med. 310:175-177, 1984. 2. C r o w l e y J. P., S h e t h, S., C a p o n e, R. J., and Sc h il l in g, R. F.: A paucity of thalassemia trait in Italian men with myocardial infarction. Acta Haemat. 78:249-251, 1987. 3. C r o w l e y, J. P., M e t z g e r, J. B., G r a y, A., M e r r il l, E., and V a l e r i, C. R.: The disposable porous bed viscometer: A new method for measuring blood viscosity. Amer. J. Clin. Path. 96:729-737, 1991. 4. F e s s a s, P., St a m a t o y a n n o p o u l o s, G., and Ke y s, A.: Serum cholesterol and thalassemia trait. Lancet 1:1182-1183, 1963. 5. G u y t o n, A. G.: Textbook of Medical Physiology, sixth edition, Philadelphia, W.B. Saunders Company, 1981, p. 207. 6. N e w m a n, W. B., et al: Relation of serum lipoprotein levels and systolic blood pressure to early atherosclerosis. New Engl. J. Med. 314: 138-144, 1986. 7. P e a r s o n, H. A. et al: Comprehensive testing for thalassemia trait. Ann. NY Acad. Sci. 232: 135-144, 1974. 8. St a s iw, D. M., et al: Some hemorheological and hematological parameters in heterozygous (3-thalassemia. Biorheology 14:1-10, 1977. 9. Str u m l a, M. M. and P h i l l i p s, M.: Effect of red cell factors on the relative viscosity of whole blood. Amer. J. Clin. Path. 39:464-474, 1963. 10. T il l m a n, W. and Sc h r o t e r, W.: Rheological properties of erythrocytes in heterozygous and homozygous p-thalassemia. Brit. J. Hemat. 43: 401-411, 1979.