COLE-COLE PARAMETERS OF HUMAN BLOOD WITH DIFFERENT

ANTICOAGULANTS

Abstract: Impedance spectroscopy of blood samples with Acid Citrate Dextrose, Ethylene Diamine Tetra-acetic
Acid, Lithium Heparin and Sodium Citrate anticaogulants, taken from 14 healthy male donors between ages 22-28,
are performed in the frequency range 20Hz-1MHz, using the two probe method. However, only high frequency data
(100kHz-1MHz) are used in fitting the Cole circle, where effects of electrode polarization are negligible. It is shown
that this is acceptable since the characteristic frequency of the samples occurs around 1 MHz. Resistivities at low
frequencies with plasma alone are calculated as 54.9 O.cm: 65.9 O.cm: 81.9 O.cm and 96.9 O.cm for EDTA, LH,
SC and ACD respectively. At 100% hematocrit, the extracellular conductivity is zero and the intracellular resistances
are 183.9 O.cm for EDTA; 180.0 O.cm for LH; 232.3 O.cm for SC and 140 O.cm for ACD. The rate of increase in
extracellular resistance with Ht is measured higher with LH and EDTA. In the physiological range of hematocrit, R
e

is the highest with LH and lowest with SC; at h=0.6 LH samples are 34% higher than SC samples. The characteristic
frequency f
c
changes with the type of anticoagulant; the frequency span at Ht=50% is from 0.8 MHz (LH) to 1.1
MHz (ACD). The angle of depression is highest with LH.


Introduction
Physiological measurements of blood are time consuming and too expensive for routine studies. On the
contrary, electrical measurements can be realized within few seconds at low cost. In a recent study, results indicated
that the complex electrical impedance measurement might provide a useful method for investigating the storage
lesions, and for tracking the quality of red cells during storage by monitoring the Cole-Cole parameters.
When applying the relaxation theory to biological tissues (Figure 1), a plot (Cole-Cole plot) of negative
reactance versus resistance of the measured impedance (Z = R + jX) is a semicircle whose center is below x=0. The
intersection of the Cole-Cole plot with the real axis provides R
0
and R

. Capacitive effects of cell membranes are

usually lumped in a constant phase angle impedance
CPA
Z
defined by:


( )
1
1/
CPA
Z K j
o
e

=
(1)




Figure 1 Equivalent electrical circuit of blood.

The depressed semicircular locus is a representation of an impedance of the form given by Equation 2 where
the angle of depression o is equal to ( ) 2 / .t o radians. The Cole-Cole equation can be written as:

( )
0
1
1
R R
Z R
j
o
et
-

= +
+
(2)
The dc resistance value R
0
is equal to R
e
. R

is given by the parallel combination of R

e
and R
i
and
( ) ( )
1
2 / .
c i e
K f R R
o
t

= +


R
e

R
i

CPA
In this study, electrical impedance spectroscopy of blood is performed over the frequency range 20 Hz-1
MHz to determine the Cole-Cole parameters, namely the resistances R

and R
o
, the characteristic frequency f
c
and
the parameter o (0 < o < 1 ). Resistivity of blood is complex and is a function of many parameters such as; the
shape of red blood cells, the volume concentration of red cells (hematocrit), the type of anticoagulant, the
temperature and the frequency. Because each donor has a different hematocrit value, either normalization is
necessary or else the results should be displayed as a function of hematocrit.
Acid Citrate Dextrose (ACD), Ethylene Diamine Tetra-acetic Acid (EDTA), Lithium Heparin (LH) and Sodium
Citrate (SC) are used as anticoagulant in the vacutainer tubes.


Materials and Method

Blood samples from 14 healthy male donors between ages 22-28 are drawn into vacuum tubes with ACD,
EDTA, LH and SC anticoagulants. From each subject; 27 ml of blood are drawn into three 9 ml Vacuettes coated
with EDTA, 30 ml blood into three 10 ml Becton Dickinson Vacutainers coated with LH, 25.5 ml blood into three
8.5 ml Becton Dickinson Vacutainers with 8,5 ml ACD-A and 27 ml blood into three SC tubes. SC tubes are
prepared in the lab by adding sodium citrate solution in a ratio of 1:9 into CAT coated 10 ml Becton Dickinson
tubes. Each donor has donated blood 4 times, in 10 days intervals. Whole blood hematocrit values for 9 donors are
given in Table 1.



Donor
#1
Donor
#2
Donor
#3
Donor
#4
Donor
#5
Donor
#6
Donor
#7
Donor
#8
Donor
#9
EDTA 48,7 51,0 44,6 48,2 48,3 45,7 49,6 50,5 51,2
LH 47,6 48,2 47,7 48,9 47,1 46,9 49,1 51,1 49,3
Sodium Citrate 44,1 46,5 43,5 45,7 44,9 - 47,7 46,2 43,2
ACD 38,9 43,8 40,5 40,2 42,5 40,0 42,0 42,3 42,0

Table 1 Whole Blood samples in different anticoagulants


Blood tubes are centrifuged at 5500 rpm for 8 minutes prior to measurements to separate red blood cells.
Cells are then diluted with plasma to generate at least three samples with hematocrit values between 30% and 60%.
Samples are measured as quickly as possible following donation.
The measurement cell is a cylindrical plastic tube of 1.45 cm diameter and 14 ml total volume, with two
stainless-steel needle electrodes of diameter 0.14 cm, separated by 1.3 cm, to inject current and measure voltage
drop across them. Complex impedance measurements are performed using the HP 4284A LCR meter in the two-
probe configuration (Figure 1).



Figure 2 The experimental set-up
Real and the imaginary parts of the impedance are measured in the frequency range 20 Hz-1 MHz. The
measurement cell constant, k = Z = 1,13 (cm
-1
) is determined with NaCl solutions of known concentrations. The
measured impedance, Z =R+ jX, is modeled with the following equations by making use of Equation 2:

( ) ( ) ( ) ( )
( ) ( ) ( )
) 1 ( 2 1
) 1 (
0
2 / . 2 1
2 / . 1
o o
o
et ot et
ot et

+ +
+
+ =
Sin
Sin R R
R R
(3)

( ) ( ) ( )
( ) ( ) ( )
) 1 ( 2 1
) 1 (
0
2 / . 2 1
2 / . .
o o
o
et ot et
ot et

+ +

=
Sin
Cos R R
X
(4)
A typical impedance spectroscopy with 2 electrodes is illustrated in Figure 3. The measured impedance is
the sum of the polarization impedance and blood sample impedance. By fitting in MatLab, Equations 3 and 4 to the
real and imaginary parts of measured impedances respectively in the frequency range from 100 kHz to 1 MHz, where
electrode polarization is negligible, the intrinsic impedance of the sample is obtained in the |-dispersion region. |-
dispersion curves predicted by Equations 3 and 4 are consistent with the measurements.
Divergence of the measured data from the calculated curves at frequencies below 10 kHz is due to electrode
polarization; o-dispersion associated with the interface between the electrodes and blood. The polarization
impedance can be obtained by subtracting this intrinsic impedance from the measured one.



Figure 3 Real and Imaginary parts of measured impedance of blood from Donor #1.
Solid lines represent plot of Equations 3 and 4 respectively, fitted to the data between 100 kHz and 10 MHz.


Effects of Changing Type of Anticogulant
Cole-Cole parameters with each anticoagulant are calculated with Matlab by applying curve fitting algorithms and
using impedance spectroscopy, 100kHz-1 MHz. These parameters are displayed in hematocrit dependent form in
Figures 4 to 7 for discussion. Extracellular and intracellular resistivities are shown to obey the Maxwell-Frick
equations, in the form (where h=Ht/100):
R
e
= a/(1-h) +b (5)
R
i
= c/h +d (6)
Results that are displayed in Figures 4 and 5 are with EDTA. The fitted equations and their corresponding correlation
coefficients are given in Table 2 for 4 types of anticoagulants.

Maxwell-Fricke Equation R
2
Re EDTA 183.7/(1-h)-128.8 0.9846
LH 199.1/(1-h)-133.2 0.9271
SC 126.9/(1-h)-45.32 0.9909
ACD 129.3/(1-h)-32.42 0.953
Ri EDTA 134/h+73.84 0.8564
LH 121.3/h+82.1 0.9215
SC 84.63/h+177.7 0.8464
ACD 134/h+24.28 0.9192

Table 2 Maxwell-Frick formulas applied to intracellular and extracellular resistances.

0.3 0.35 0.4 0.45 0.5 0.55 0.6
140
160
180
200
220
240
260
280
300
320
340
h
R
e
=
E
x
t
r
a
c
e
l
l
u
l
a
r

R
e
s
i
s
t
a
n
c
e
(
o
h
m
)
EDTA


Re vs. h
M-F:183.7/(1-h)-128.8 R=0.9846

Figure 4 Extracellular resistance of blood with EDTA, as a function of h. Mean data from 9 donors are considered.

0.3 0.35 0.4 0.45 0.5 0.55 0.6
300
350
400
450
500
h
R
i
=
I
n
t
r
a
c
e
l
l
u
l
a
r

R
e
s
i
s
t
a
n
c
e
(
o
h
m
)
EDTA


Ri vs. h
M-F:134/h+73.84 R=0.8564

Figure 5 Intracellular resistance of blood with EDTA, as a function of h. Mean data from 9 donors are considered.

Equation 5 represents the Maxwell-Frick formula which is valid at low frequencies when cells do not
conduct. When comparing the equations in Table 2, the rate of increase in extracellular resistance with Ht is
measured higher with LH and EDTA. In the physiological range of hematocrit, R
e
is the highest with LH and lowest
with SC; at h=0.6 LH samples are 34% higher than SC samples. Amounts of anticoagulants in the vacutainer tubes
influence the extracellular resistance. The fact that each type of anticoagulant has a different resistivity must also be
taken into consideration. In Zhao paper the resistivities were reported as 60.5 O.cm for ACD: 20.6 O.cm for EDTA
and 53.4 O.cm for SC, at 37
0
C. In this study, the ratios of anticogulants to total sample volumes are 15% ACD,
1.2% EDTA, 9.9% SC and % LH. Therefore, resistivities of blood samples with ACD and SC anticoagulants are
expected to be influenced the most.


Figure 6 Cole-Cole parameter o of blood with different anticoagulants, as a function of hematocrit.
The resistivities at low frequencies when h=0, i.e. in case blood cells are completely removed, are calculated
as 54.9 O.cm: 65.9 O.cm: 81.9 O.cm and 96.9 O.cm for EDTA, LH, SC and ACD respectively. At 100% hematocrit
the extracellular conductivity is zero. From Table 2, the intracellular resistances are 207.8 O for EDTA; 203.4 O for
LH; 262.3 O for SC and 158.3 O for ACD. In previous results of Pauly and Schwan 1977; Wojcik and Lambotte
1977 and Zhao 1993, the intracellular resistivity was reported as 194 O.cm at 25
0
C.




Figure 7 Characteristic frequency of blood with different anticoagulants.


The characteristic frequency f
c
changes with the type of anticoagulant; the frequency span at Ht=50% is
from 0.8 MHz (Lithium Heparin) to 1.1 MHz (Acid Citrate Dextrose). f
c
gets lower for each anticoagulant as the
volume occupied by red blood cells become greater implying a larger effective cell membrane capacitance. The
effect is more pronounced with LH.


Conclusion

Both plasma and eryhtrocyte electrical properties are affected with anticoagulants and the effect is traceable
in the Cole-Cole parameters. All measurements are displayed as a function of eryhtyocyte volume ratio in blood
samples. Resistivities at low frequencies with plasma alone vary from 54.9 O.cm (EDTA) to 96.9 O.cm (ACD).
When h=1, the intracellular resistance is the highest with SC (232.3 O.cm) and the lowest with ACD (140 O.cm)
Extracellular resistance increases with Ht. In the physiological range of 60%Ht, R
e
is the highest with LH and lowest
with SC. The characteristic frequency f
c
changes with the type of anticoagulant; the frequency span at Ht=50% is
from 0.8 MHz (LH) to 1.1 MHz (ACD). The angle of depression is highest with LH.

References
1) Zhao, T.X., Electrical Impedance and Haematocrit of Human Blood with Various
Anticoagulants, Physiological Measurement, Vol. 14, pp. 299-307, 1993.
2) Cha, K., R.G. Faris , E.F. Brown, and D.W. Wilmore, An Electronic Method for
Rapid Measurement of Haematocrit in Blood Samples, Physiological
Measurement, Vol. 15, pp. 129-137, 1994.
3)Zhao, T.X., B. Jacobson, and T. Ribbe, Triple Frequency Method for Measuring
Blood Impedance," Physiological Measurement, Vol. 14, pp. 145-156, 1993.
4)Schwan,H.P.,and C.F Kay, The Conductivity of Living Tissues,Ann.NY Acad.Sci.,
Vol 65,pp.1007-1013,1956
5) Cole, K.S., Membranes, Ions, and Impulse, University of California Press,
Berkeley, 1972.
6) Hill, D.W., and F.D. Thompson, The Effect of Hematocrit on the Resistivity of
Human Blood at 370C and 100 kHz, Med. Biol. Eng., Vol. 13, pp. 182-186, 1975.
7) Mohapatra, S.N., and D.W. Hill, The Changes in Blood Resistivity with
Hematocrit and Temperature, Eur. J. Intensive Care Med., Vol. 1, pp. 153-162,
1975.
8) Sezdi, M., Electrical Impedance Spectroscopy of Human Blood, MS. Thesis,
Boazii University, 1998.
9) Fricke, H., A Mathematical Treatment of the Electric Conductivity and Capacity
of Disperse Systems, Phys. Rev., Vol. 24, pp. 575-587, 1924.