White blood cell and platelet counts could affect whole blood viscosity

Abstract

Blood viscosity is correlated with cerebral blood flow and cardiac output, and increased viscosity may increase the risk of thrombosis or thromboembolic events. The relationship between hematocrit and viscosity is well-known, however, the relationships between white blood cell (WBC) or platelet count and viscosity were not fully studied. The aim of the present study was to determine the influences of platelet count and WBC count on blood viscosity.
One-hundred and 13 subjects with different hemoglobin, WBC and platelet count were enrolled into the study. The variables measured included serum fibrinogen, cholesterol, triglyceride, high-density lipoprotein (HDL), low-density lipoprotein (LDL), complete blood counts including hemoglobin, hematocrit, platelet count, red blood cell (RBC) count, WBC count, whole blood and plasma viscosity. The relationships of these variables with whole blood or plasma viscosity were analyzed.
Serum fibrinogen, cholesterol, triglyceride, HDL and LDL did not correlate with whole blood viscosity. Not only hematocrit, hemoglobin and RBC, but also WBC and platelet count, could affect whole blood viscosity. On the other hand, none of the variables could affect plasma viscosity.
All the blood cell components, but not the plasma proteins detected above, could affect whole blood viscosity. When patients are with high leukocytosis and thrombocytosis, impaired blood viscosity should also be considered to obtain appropriate clinical management.

Full text

B
lood viscosity is correlated with cerebral blood
flow
1
and cardiac output,
2
and increased viscosity
may increase the risk of thrombosis or thromboembolic
events.
3
Hematocrit and hemoglobin are among the
many factors that influence blood viscosity. Other fac
-
tors such as white blood cell (WBC) count and platelet
count are less evaluated. We hypothesized that WBC and
platelet count might also affect blood viscosity. In the
present study, we planned to evaluate the influence of
different levels of red blood cell (RBC), hematocrit, he
-
moglobin, WBC and platelet count on blood viscosity.
METHODS
One-hundred and 13 patients who visited the hema
-
tology clinic with different hemoglobin, WBC and platelet
counts were enrolled into our study. Iron deficiency ane-
mia, pancytopenia, polycythemia vera, essential throm-
bocythemia, idiopathic thrombocytopenic purpura, mye-
lodysplastic syndrome, aplastic anemia and thalassemia
composed most of the disorders. All patients with multi
-
ple myeloma, Waldenstrom macroglobulinemia and cir
-
rhosis of liver with reverse albumin/globulin ratio were
excluded as they would have high plasma viscosity and
might affect our results. After having signed informed con
-
sent, these patients were measured for serum fibrinogen,
cholesterol, triglyceride, high-density lipoprotein (HDL),
low-density lipoprotein (LDL), complete blood counts
(CBC), whole blood and plasma viscosity. Whole blood
and plasma viscosity were measured by a digital vis
-
cometer (Brookfield Engineering Lab, Stoughton, MA,
USA). CBC were measured by an automated hematology
analyzer (Sysmex SE9000, Toa Medical Electronics,
394
J Chin Med Assoc
2004;67:394-397
Chao-Hung Ho
Division of Hematology, Department of
Internal Medicine, Taipei Veterans
General Hospital; National Yang-Ming
University School of Medicine, Taipei,
Taiwan, R.O.C.
Key Words
hematocrit;
plasma viscosity;
platelet count;
white blood cell count;
whole blood viscosity
Original Article
White Blood Cell and Platelet Counts
Could Affect Whole Blood Viscosity
Background. Blood viscosity is correlated with cerebral blood flow and cardiac out
-
put, and increased viscosity may increase the risk of thrombosis or thromboembolic
events. The relationship between hematocrit and viscosity is well-known, however,
the relationships between white blood cell (WBC) or platelet count and viscosity
were not fully studied. The aim of the present study was to determine the influences
of platelet count and WBC count on blood viscosity.
Methods. One-hundred and 13 subjects with different hemoglobin, WBC and platelet
count were enrolled into the study. The variables measured included serum
fibrinogen, cholesterol, triglyceride, high-density lipoprotein (HDL), low-density li
-
poprotein (LDL), complete blood counts including hemoglobin, hematocrit, platelet
count, red blood cell (RBC) count, WBC count, whole blood and plasma viscosity.
The relationships of these variables with whole blood or plasma viscosity were ana
-
lyzed.
Results. Serum fibrinogen, cholesterol, triglyceride, HDL and LDL did not correlate
with whole blood viscosity. Not only hematocrit, hemoglobin and RBC, but also
WBC and platelet count, could affect whole blood viscosity. On the other hand, none
of the variables could affect plasma viscosity.
Conclusions. All the blood cell components, but not the plasma proteins detected
above, could affect whole blood viscosity. When patients are with high leukocytosis
and thrombocytosis, impaired blood viscosity should also be considered to obtain ap-
propriate clinical management.
Received: February 13, 2004.
Accepted: July 1, 2004.
Correspondence to: Chao-Hung Ho, MD, Division of Hematology, Taipei Veterans General Hospital,
201, Sec. 2, Shih-Pai Road, Taipei 112, Taiwan.
Tel and Fax: +886-2-2875-7106; E-mail: chho@vghtpe.gov.tw

Kobe, Japan); serum cholesterol, triglyceride, HDL,
LDL, were determined using the SMAC-II Analyzer
(Technicon, New York, USA). Kruskal-Wallis test was
used to analyze the results. Pearson correlation matrix
was used to assess the relationship between the values.
RESULTS
Table 1 shows the mean and range of different vari
-
ables in the 113 subjects. Figs. 1 and 2 show the relation
-
ships of whole blood viscosity with RBC, hemoglobin,
hematocrit, log-WBC and log-platelet count in these sub
-
jects. These figures demonstrate that there was significant
relationship between whole blood viscosity and RBC, he
-
moglobin, hematocrit, log-WBC and log-platelet count
(p < 0.001, < 0.001, < 0.001, = 0.005 and 0.016, respec
-
tively).
We then divided the patients into 3 subgroups ac
-
cording to their levels of whole blood viscosity in order
to see the difference in each variable. Group A included
58 patients with whole blood viscosity £ 3.0, Group B in
-
cluded 39 patients with whole blood viscosity 3.01-4.0
and Group C, 16 patients with whole blood viscosity >
4.0. Table 2 shows the means of the variables in these 3
subgroups. Hemoglobin, hematocrit, RBC, WBC and
platelet counts were significantly different in the 3 sub
-
groups of patients with different whole blood viscosity
395
August 2004 WBC, Platelet Count and Viscosity
RBC (M/UL)
012345678
0
1
2
3
4
5
6
7
r=0.639
p<0.001
Hb (g/dl)
0 5 10 15 20
WBV (cps)
0
1
2
3
4
5
6
7
r=0.695
p<0.001
Hct (%)
0 10203040506070
WBV (cps)
0
1
2
3
4
5
6
7
r=0.706
p<0.001
WBV = Whole blood viscosity
Fig. 1. The relationship between RBC, hemoglobin, hematocrit and whole blood viscosity.
Table 1. Mean of different variables in 113 subjects
Male Female Total
Mean ± SD (Range)
Age (year) 72.0 ± 11.6 (25-87) 57.4 ± 18.7 (15-89) 66.3 ± 16.3 (15-89)
Hemoglobin (g/dL) 10.5 ± 3.4 (5.2-19.2) 9.1 ± 3.0 (2.9-14.7) 9.9 ± 3.3 (2.9-19.2)
Hematocrit (%) 32.2 ± 10.4 (14.7-59.1) 28.3 ± 8.0 (10.9-44.3) 30.7 ± 9.7 (10.9-59.1)
Red cell count (´ 10
12
/L) 3.70 ± 1.31 (1.52-6.89) 3.71 ± 1.09 (1.79-6.69) 3.70 ± 1.22 (1.52-6.89)
MCV (fL) 89.0 ± 11.2 (63.4-121.1) 78.0 ± 14.2 (52.1-102.5) 84.7 ± 13.5 (52.1-121.1)
White cell count (´ 10
9
/L) 6.4 ± 6.7 (0.91-46.4) 13.9 ± 49.6 (1.77-334.6) 9.3 ± 31.4 (0.91-334.6)
Platelet count (´ 10
9
/L) 202 ± 200 (1-784) 292 ± 353 (3-1935) 237 ± 272 (1-1935)
Fibrinogen (mg/dL) 382 ± 114 (190-650) 342 ± 88 (106-501) 367 ± 272 (106-650)
Cholesterol (mg/dL) 160 ± 45 (50-254) 173 ± 49 (91-313) 166 ± 47 (50-313)
Triglyceride (mg/dL) 126 ± 121 (29-941) 113 ± 50 (33-234) 121 ± 99 (29-941)
Whole blood viscosity 3.16 ± 0.88 (1.53-6.05) 2.81 ± 0.60 (1.69-4.21) 3.03 ± 0.80 (1.53-6.05)
Plasma viscosity 1.31 ± 0.28 (0.40-1.99) 1.27 ± 0.18 (0.95-1.85) 1.30 ± 0.24 (0.40-1.99)

(p < 0.005, < 0.005, < 0.005, = 0.005 and 0.016, respec
-
tively, Kruskal-Wallis test). On the other hand, MCV,
fibrinogen, cholsterol, triglyceride, HDL and LDL (data
of the latter 2 not shown) were not significantly different
among them. Thus, whole blood viscosity was affected
significantly by hemoglobin, hematocrit, RBC and WBC
counts, but unaffected by fibrinogen, cholesterol, tri
-
glyceride and platelets.
We further divided the patients into 3 subgroups ac
-
cording to their levels of plasma viscosity. Group D in
-
cluded 41 patients with plasma viscosity £ 1.2, Group E
included 41 patients with plasma viscosity 1.2-1.4 and
Group F, 26 patients with plasma viscosity > 1.4. No sig
-
nificant difference was found in any of the variables in
these 3 subgroups (p all > 0.05). Thus, no variables af
-
fected plasma viscosity.
DISCUSSION
Blood viscosity is known to have close relationship
with blood flow. Increased viscosity may increase the
risk of thrombosis or thromboembolic events. Blood vis
-
cosity can be divided into whole blood viscosity and
plasma viscosity. These 2 viscosities can exert different
influences in different situations. For example, in the ce
-
rebral area, blood flow is more affected by plasma vis
-
cosity rather than whole blood viscosity.
1,4
High whole
blood viscosity with high hemoglobin concentration
might increase the risk of poor pregnancy outcome,
5
or
the potential for increased thrombosis.
6
Many factors
can affect blood viscosity, such as red blood cell count,
hematocrit, immunoglobulins and fibrinogen. The aim
of the present study focused on the blood cells which
Chao-Hung Ho Journal of the Chinese Medical Association Vol. 67, No. 8
396
Fig. 2. The relationship between log-WBC, log-platelet and whole blood viscosity.
Table 2. Mean of various variables in 3 subgroups with different levels of whole blood viscosity
Group A (£ 3, 58)
a
Group B (3.01-4, 39)
a
Group C (> 4, 16)
a
p value
Mean ± SD
Age (year) 66.1 ± 15.6 66.7 ± 18.5 67.8 ± 12.4 0.868
Hemoglobin (g/dL) 8.2 ± 2.1 11.1 ± 2.6 13.6 ± 4.3 < 0.005
Hematocrit (%) 25.6 ± 5.9 34.0 ± 7.1 42.3 ± 13.9 < 0.005
Red cell count (´ 10
12
/L) 3.07 ± 0.84 4.17 ± 0.93 4.89 ± 1.72 < 0.005
MCV (fL) 85.4 ± 15.1 82.4 ± 12.7 87.5 ± 7.6 0.734
White cell count (´ 10
9
/L) 10.5 ± 43.4 6.2 ± 3.7 12.8 ± 12.0 0.005
Platelet count (´ 10
9
/L) 170 ± 164 303 ± 376 323 ± 246 0.016
Fibrinogen (mg/dL) 373 ± 101 366 ± 101 341 ± 151 0.896
Cholesterol (mg/dL) 159 ± 49 176 ± 40 159 ± 58 0.190
Triglyceride (mg/dL) 123 ± 127 114 ± 56 128 ± 57 0.270
a
Whole blood viscosity & number of cases.

would affect blood viscosity.
In our study, the levels of hemoglobin, hematocrit,
RBC, WBC and platelet counts were in a rather great
range as the patients were from the hematology clinic.
Thus, we had a chance to check the influence of high
platelet and high white blood cell counts on viscosity.
In our study, we used 2 methods of classification to
detect the differences among variables. One was to di
-
vide the patients according to their levels of viscosity to
see the differences of every variable, and the other was to
show the relationship between the viscosity and different
variables. Both methods showed that no variables could
affect the plasma viscosity, including fibrinogen, choles
-
terol, triglyceride, HDL and LDL as well as all the blood
cell components. On the other hand, hemoglobin, hemat
-
ocrit, RBC and WBC counts were found to affect the
whole blood viscosity, whichever method was used.
RBC and their related indexes are well known to af
-
fect viscosity. We further proved this fact. Though their
influence found in the present in vitro study was mainly
on whole blood viscosity rather than on plasma viscosity,
in microcirculation, increased hemoglobin would de-
crease the deformability and thus increase plasma vis-
cosity.
4
This should be kept in mind when the blood flow
occurs in vivo.
Since high leukocytosis and high platelet count can-
not be easily found in medical settings other than hema-
tology, their influences on viscosity have seldom been
investigated. Our study gave us a chance to know their
influence on viscosity and, thus, that they might also be a
possible risk factor of thromboembolic disease. Further
study should be performed to make more solid conclu
-
sions.
In conclusion, not only the RBC but also the WBC
and platelet count can affect whole blood viscosity. Fur
-
ther studies should be performed to identify their clinical
significance.
ACKNOWLEDGEMENTS
This research was supported by a grant (VGH 91-53)
from Veterans General Hospital, Taipei, Taiwan, R.O.C.
REFERENCES
1. Tomiyama Y, Brian JE Jr, Todd MM. Plasma viscosity and ce
-
rebral blood flow. Am J Physiol 2000;279:1949-54.
2. Metivier F, Marchais SJ, Guerin AP, Pannier B, London GM.
Pathophysiology of anaemia: focus on the heart and blood
vessels. Nephrol Dialy Transp 2000;15(suppl):14:1-8.
3. Kwaan HC, Bongu A. The hyperviscosity syndromes. Semin
Thromb Hemostas 1999;25:199-208.
4. Metry G, Wikstrom B, Valind S, Sandhagen B, Linde T,
Beshara S, et al. Effect of normalization of hematocrit on
brain circulation and metabolism in hemodialysis patients. J
Am Soc Nephrol 1999;10:854-63.
5. Yip R. Significance of an abnormally low or high hemoglobin
concentration during pregnancy: special consideration of iron
nutrition. Am J Clin Nutr 2000;72(suppl):272-9.
6. Shaskey DJ, Green GA. Sports haematology. Sports Med
2000;29:27-38.
397
August 2004 WBC, Platelet Count and Viscosity