# Bioimpedance in the assessment of unilateral lymphedema of a limb: the optimal frequency.

**ABSTRACT** Bioimpedance techniques provide a reliable method of assessing unilateral lymphedema in a clinical setting. Bioimpedance devices are traditionally used to assess body composition at a current frequency of 50 kHz. However, these devices are not transferable to the assessment of lymphedema, as the sensitivity of measuring the impedance of extracellular fluid is frequency dependent. It has previously been shown that the best frequency to detect extracellular fluid is 0 kHz (or DC). However, measurement at this frequency is not possible in practice due to the high skin impedance at DC, and an estimate is usually determined from low frequency measurements. This study investigated the efficacy of various low frequency ranges for the detection of lymphedema.

Limb impedance was measured at 256 frequencies between 3 kHz and 1000 kHz for a sample control population, arm lymphedema population, and leg lymphedema population. Limb impedance was measured using the ImpediMed SFB7 and ImpediMed L-Dex(®) U400 with equipotential electrode placement on the wrists and ankles. The contralateral limb impedance ratio for arms and legs was used to calculate a lymphedema index (L-Dex) at each measurement frequency. The standard deviation of the limb impedance ratio in a healthy control population has been shown to increase with frequency for both the arm and leg. Box and whisker plots of the spread of the control and lymphedema populations show that there exists good differentiation between the arm and leg L-Dex measured for lymphedema subjects and the arm and leg L-Dex measured for control subjects up to a frequency of about 30 kHz.

It can be concluded that impedance measurements above a frequency of 30 kHz decrease sensitivity to extracellular fluid and are not reliable for early detection of lymphedema.

**2**Bookmarks

**·**

**138**Views

- [Show abstract] [Hide abstract]

**ABSTRACT:**The successful application of impedance spectroscopy in daily practice requires accurate measurements for modeling complex physiological or electrochemical phenomena in a single frequency or several frequencies at different (or simultaneous) time instants. Nowadays, two approaches are possible for frequency domain impedance spectroscopy measurements: (1) using the classical technique of frequency sweep and (2) using (non-)periodic broadband signals, i.e. multisine excitations. Both techniques share the common problem of how to design the experimental conditions, e.g. the excitation power spectrum, in order to achieve accuracy of maximum impedance model parameters from the impedance data modeling process. The original contribution of this paper is the calculation and design of the D -optimal multisine excitation power spectrum for measuring impedance systems modeled as 2R-1C equivalent electrical circuits. The extension of the results presented for more complex impedance models is also discussed. The influence of the multisine power spectrum on the accuracy of the impedance model parameters is analyzed based on the Fisher information matrix. Furthermore, the optimal measuring frequency range is given based on the properties of the covariance matrix. Finally, simulations and experimental results are provided to validate the theoretical aspects presented.Measurement Science and Technology 01/2012; 23(8):085702. · 1.44 Impact Factor

Page 1

Bioimpedance in the Assessment of Unilateral

Lymphedema of a Limb: The Optimal Frequency

Richelle Gaw, Ph.D.,1Robyn Box, M.Phty., Ph.D.,2and Bruce Cornish, Ph.D.1

Abstract

Background: Bioimpedance techniques provide a reliable method of assessing unilateral lymphedema in a

clinical setting. Bioimpedance devices are traditionally used to assess body composition at a current frequency of

50kHz. However, these devices are not transferable to the assessment of lymphedema, as the sensitivity of

measuring the impedance of extracellular fluid is frequency dependent. It has previously been shown that the

best frequency to detect extracellular fluid is 0kHz (or DC). However, measurement at this frequency is not

possible in practice due to the high skin impedance at DC, and an estimate is usually determined from low

frequency measurements. This study investigated the efficacy of various low frequency ranges for the detection

of lymphedema.

Methods and Results: Limb impedance was measured at 256 frequencies between 3kHz and 1000kHz for a

sample control population, arm lymphedema population, and leg lymphedema population. Limb impedance

was measured using the ImpediMed SFB7 and ImpediMed L-Dex?U400 with equipotential electrode placement

on the wrists and ankles. The contralateral limb impedance ratio for arms and legs was used to calculate a

lymphedema index (L-Dex) at each measurement frequency. The standard deviation of the limb impedance ratio

in a healthy control population has been shown to increase with frequency for both the arm and leg. Box and

whisker plots of the spread of the control and lymphedema populations show that there exists good differen-

tiation between the arm and leg L-Dex measured for lymphedema subjects and the arm and leg L-Dex measured

for control subjects up to a frequency of about 30kHz.

Conclusions: It can be concluded that impedance measurements above a frequency of 30kHz decrease sensi-

tivity to extracellular fluid and are not reliable for early detection of lymphedema.

Introduction

B

lymphedema. Lymphedemais characterized byanincrease in

extracellular fluid. Traditional methods of assessment include

circumferential measurements and volume by calculation,

perometry, or Archimedes principle. However, each method

measures only the total fluid volume of the limb and is not

sensitive to changes in extracellular fluid. Bioimpedance

techniques (both single frequency and bioimpedance spec-

troscopy, BIS) can measure changes in extracellular fluid and

have been used to assess unilateral lymphedema.1,2In this

case, the variation from ‘‘normal’’ of the ratio ofthe affected or

at risk limb impedance to the healthy contralateral limb im-

pedance is used to assess the degree of lymphedema present.

ioimpedance devices are rapidly gaining momentum

as an adjunct in the early detection of unilateral limb

Bioimpedance devices measure the electrical impedance of

biological tissue in response to an applied alternating current.

The electric current will pass through different tissues de-

pendent upon the impedance to flow. Electric currents ap-

pliedtothebodyareprimarilydistributedinfluidsandblood,

due to the low resistivity of these tissues. In the presence of

lymphedema, the applied current will travel predominantly

through the accumulation of lymphatic fluids. This noninva-

sive and cost-effective technology can be used as a reliable

tool to assess the presence of lymphedema in patients with

either upper or lower limb affliction at an early stage.3

The electrical properties of tissue can be described by elec-

trically similar components. When current is applied to the

tissue, the extra- and intracellular fluids behave as resis-

tive components, while the cell membranes behave as reactive

components.At0kHz(directcurrentorDC)thecellmembrane

1Faculty of Science and Technology, Queensland University of Technology, Brisbane, Australia.

2QLD Lymphoedema and Breast Oncology Physiotherapy, Brisbane, Australia.

The research studies have been funded by in kind support from ImpediMed Ltd, Brisbane, Australia, with clinical measurements of test

subjects obtained independently by Dr. R. Box.

LYMPHATIC RESEARCH AND BIOLOGY

Volume 9, Number 2, 2011

ª Mary Ann Liebert, Inc.

DOI: 10.1089/lrb.2010.0020

93

Page 2

acts as an insulator and all current flows through the extra-

cellular fluid. The impedance measured at DC (R0) is that of

extracellular fluid only. At higher current frequencies (alter-

nating current or AC), the cell membrane becomes a con-

ductor and the current will flow in both the extracellular and

intracellular fluid in a ratio that is dependent of the applied

current frequency. The impedance measured at a single fre-

quency is therefore a combination of the contribution from

extra-andintracellularfluid.Asthefrequencyincreases,there

will become a practical frequency at which the measured

impedance is no longer sensitive enough to track extracellular

fluid alone.

Due to practical limitations, R0cannot be measured di-

rectly. However, knowledge of the frequency response of bi-

ologicalmaterialallows finite

extrapolated to derive this value. Bioimpedance spectroscopy

(BIS) devices that use up to 256 frequency measurements to

extrapolate to the DC value have been shown to be more

sensitive to extracellular fluid changes.4It has been shown

that R0is most sensitive to changes in extracellular fluid5and

is therefore the best index of lymphedema.

Ithasbeensuggestedthatsinglefrequencyimpedancelimb

ratios and R0limb ratios are essentially interchangeable in

assessing lymphedema as long as the current frequency re-

mains low.6However, the same study also reports an increase

in the mean difference between the R0and single frequency

methods and a widening of the 95% confidence interval of the

bias as the measurement frequency of the device increased.

The study does not address a specific cut-off frequency for

accurate assessment of lymphedema. The confidence interval

increases from ?1.2% to 2.8% at 10kHz to ?4% to 7.1% at

50kHz, approximately 250% increase in interval width. This

suggests that lymphedema assessments using a current fre-

quency of 50kHz do not have adequate sensitivity. The study

also highlights the need for appropriate detection ratios for

leg lymphedema. In another study, a frequency range of

5–10kHz has been identified as the optimum for single fre-

quency measurement of lower leg swelling due to blood

pooling.7

Many single frequency bioimpedance devices (predomi-

nantly 50kHz) are available on the market. These devices are

traditionally used to determine body composition parame-

ters, including total body water and extracellular water of the

whole body. However, at 50kHz, the current passes through

both intra- and extracellular fluid and so the determination of

the fluid levels are not dependant on extracellular fluid alone.

Body composition devices are also designed in most cases

to measure hand to foot impedance to determine whole body

composition using prediction equations based on population

studies. The hand to foot measurement consists of the addi-

tion of the arm, trunk, and leg impedance. Whole body water

content (measured using specific total body water apparatus

measured at 50kHz) has been shown to be insensitive to de-

creases in circumferential and volumetric measures of an arm

during the treatment of lymphedema.8Therefore, changes in

extracellular fluid volume in a single limb would be better

identified by a single limb measurement. Bioimpedance has

been shown to reliably measure specific body segments (arms

and legs) when the measurement electrodes are positioned on

the contralateral limb, thus using the concept of electrical

equipotentials. This approach enables accurate and repro-

ducible results.9

frequencydatatobe

Therefore the aim of this study is to demonstrate the fre-

quency dependence of the precision of unilateral lymphede-

ma assessment using bioimpedance techniques and to

investigate the efficacy of various frequency ranges for the

accurate discrimination between lymphedema and control

subjects. This will lead to a more informed and reliable choice

in the selection of current frequency for the assessment of

lymphedema.

Materials and Methods

In order to determine the dependence of the assessment of

unilateral lymphedema on the current frequency used to

measure the bioelectrical impedance, the impedance of each

limb was measured at 256 frequencies. This was completed

for a control population and populations with clinically evi-

dent unilateral lymphedema in either the arm or the leg,

classified according to a criterion referenced scale modified

from that described by Miller et al.10The data used in this

study was collected as part of three research studies con-

ducted with the Queensland University of Technology (QUT)

and the University of Queensland (UQ). Ethics approval was

obtained from the respective authoritative bodies and in-

formed consent was obtained from subjects prior to partici-

pation. Each data subset was used for analysis of one of the

three sample population groups. These were the control

population, unilateral arm lymphedema population, and

unilateral leg lymphedema population.

The limb impedance of each participant was measured

using either the ImpediMed L-Dex U400 or the ImpediMed

SFB7, according to standard measurement procedures. The L-

Dex U400 instrumentation implements the same BIS platform

utilized in the SFB7 to measure the raw bioimpedance data,

thus generating equivalent raw impedance values. The dif-

ferences between the two devices are aesthetic where the

software and user interface of the L-Dex U400 have been

specifically designed for use in assessing lymphedema in a

clinical environment. Both devices measure the impedance at

256 frequencies from 3kHz to 1000kHz and perform regres-

sion analysis to determine R0and R?(the theoretical resis-

tance at infinite frequency).

Limb impedance was measured using an equipotential

electrode placement on the wrists and ankles.9The limb im-

pedance ratio was calculated for all 256 current frequencies.

For the control population, the nondominant to dominant

limb ratio was calculated. For the test populations, the unaf-

fected to affected limb ratio was calculated. The limb im-

pedance ratio was then used to determine a lymphedema

index (L-Dex), used to assess the presence of lymphedema. In

the case of SFB7 measurements, the raw impedance was

converted to an L-Dex score using proprietary software sup-

plied by the manufacturer.

The L-Dex is a quantitative indicator that represents the

comparison of the measured limb impedance ratio to a ‘nor-

mal’ range of limb impedance ratios established in a healthy

population matched for limb dominance and gender. This is

unlike body composition algorithms that use prediction

equations to determine the amount of extracellular water

content from impedance measurements. The simple lineari-

zation of the normal distribution means that measured im-

pedance ratios equal to the mean ratio of the equivalent

healthy population will have an L-Dex of 0 and those greater

94GAW ET AL.

Page 3

than three standard deviations from the equivalent healthy

population range will have an L-Dex greater than 10. The use

of the L-Dex for the assessment of lymphedema has been

previously reported.11,12These studies identify the need for a

simple diagnostic scale to be used in the assessment of lym-

phedema. The use of L-Dex achieves this and allows the re-

sults to be directly compared across gender and dominance of

the affected limb on one simple scale. This avoids the need for

separate analysis to be performed for different population

characteristics (eg, female dominant arm affected) that is re-

quired when working with raw impedance ratios.

Arm lymphedema population

The sample unilateral arm lymphedema population data

was collected at the Queensland Lymphedema and Breast

Oncology Physiotherapy Clinic (UQ ethics approval No.

2007001013). Data was collected from all eligible and con-

senting patients previously diagnosed with arm lymphedema

who attended the clinic during the study duration. Data col-

lection for patients with mild to severe unilateral arm lym-

phedema secondary to breast cancer surgery was conducted

using the ImpediMed L-Dex U400 and current clinical prac-

tice to create a dataset indicative of a sample arm lymphe-

dema test population. The demographics of the arm

lymphedema population are shown in Table 1. The imped-

ance of both arms was measured at 256 current injection fre-

quencies and the unaffected/affected arm impedance ratio

was calculated and the L-Dex recorded. The clinical charac-

teristics of the lymphedema were graded according to the

modified scale from that of Miller et al.10and the ImpediMed

XCA. Of the 15 subjects recruited into this population group,

only data from 12 subjects were analyzed, based on the cri-

teria that at least two of the three clinical assessment methods

must confirm the presence of lymphedema at the time of as-

sessment.

Leg lymphedema population

The sample unilateral leg lymphedema population was

collected at the Queensland Lymphedema and Breast On-

cology Physiotherapy Clinic (UQ ethics approval No.

2008001030). Data was collected from all eligible and con-

senting patients previously diagnosed with leg lymphedema

who attended the clinic during the study duration. Data col-

lection for patients with mild to severe unilateral leg lym-

phedema was conducted using the ImpediMed L-Dex U400

and current clinical practice to create a dataset indicative of a

sample leg lymphedema test population. The demographics

of the leg lymphedema population are shown in Table 1. The

impedance of both legs was measured at 256 current fre-

quenciesandtheunaffected/affectedlegimpedanceratiowas

calculated and the L-Dex recorded. The clinical characteristics

of the lymphedema were graded according to the scale

modified from that of Miller et al.10and circumferential vol-

ume differences using a tape measure. Of the 16 subjects re-

cruited into this population group, only the data from 12

subjectswasanalysed, basedonthecriteriathatatleast two of

the three clinical assessment methods must confirm the

presence of lymphedema at the time of data collection.

Control population

The control population data was collected at the Queens-

land University of Technology (QUT ethics approval No.

0700000853). A survey of the healthy population was con-

ducted using the ImpediMed SFB7 for the limb impedance

ratiosat256current frequencies between3kHzand 1000kHz.

Sixty-five self-diagnosed healthy females who met eligibility

criteria were recruited from the staff and student population.

The demographics of the control population are shown in

Table 1. The impedance of both arms and legs was measured

and the nondominant/dominant arm and leg impedance ra-

tio and L-Dex was calculated using the proprietary software

supplied by the manufacturer

Results and Discussion

The mean subject age was significantly higher in the test

populations than the control population (p<0.0001). The

difference of the mean height between each population and

the mean weight between each population was not significant

(arms: p¼0.1874, legs: p¼0.9453, and arms: p¼0.4472, legs:

p¼0.6163 for height and weight, respectively).

The arm and leg impedance ratios of the control and test

populations were calculated for each frequency between

3kHz and 1000kHz. The mean and standard deviation of the

impedanceratiowasalsocalculatedateachfrequency,aswell

as R0and R?. The variation of the standard deviation for the

arm and leg impedance ratios of the control population as a

function of current measurement frequency is shown in Fig-

ure 1. The standard deviation of the arm impedance ratio at

0kHz is 0.031. This is comparable to other published exam-

ples that report the standard deviation of female arm ratios to

be 0.034.13It is hypothesized that the lower standard devia-

tion in the present study is due to advancements in instru-

mentation technology that produces more reliable and

repeatable measurements. The 25th–75thinterquartile range of

the nondominant to dominant limb impedance ratio of a

healthy control subject has also been reported as 0.994–1.057

Table 1. Population Demographics for Control, Arm Lymphedema, and Leg Lymphedema Subjects

Parameter (m?s)

No. of Subjects

Age (years)

Height(cm)

Weight (kg)

Type of lymphedema

Modified Millers Grading

L-Dex Score

Control GroupArm Lymphoedema GroupLeg Lymphoedema Group

65 1212

40.8?11.6

165.1?6.3

70.5?14.6

N/A

4?0

0?3.3

61.2?9.2

162.7?3.1

74.0?13.8

12 Secondary

10.3?2.2

44.0?37.8

60.5?6.5

165.0?7.1

72.7?11.5

9?1.7

24.8?14.4

9 Secondary/3 Primary

LYMPHEDEMA ASSESSMENT: OPTIMAL IMPEDANCE FREQUENCY95

Page 4

(0.063).2This suggests a 50% distribution (in comparison to

68% for one standard deviation) of?0.0315, which is com-

parable to the standard deviation found by this current study.

There has been no standard deviation previously published

for healthy leg impedance ratios.

The standard deviation for the leg ratio is consistently

higher than for the arm ratio. This is due to better accuracy in

measuring an arm. In the case of the leg, the geometry of the

body means that often the legs cannot be separated enough to

provide adequate insulation between the thighs. As a result,

impedance measurements of the leg may contain more noise

than the arm due to this added interference.

Figure 1 clearly shows that the standard deviation of the

limb impedance ratio of a normal healthy population in-

creases as the current frequency increases for both the arm

and the leg. This is more evident in the standard deviation of

the arm ratios which increases from 3kHz. The standard de-

viation of the leg ratios begins to increase more rapidly at

20kHz. These results indicate that the natural spread of the

measured impedance ratio within a normal population in-

creases with the measurement frequency.

The variance of the limb impedance ratio is expected to be

smallest at 0kHz; however, this is not the case for the legs. In

the legs, the variance decreases from 0.047at 0kHz to 0.046 at

20kHz. Again, this may be due to inherent noise in a leg

measurement induced by the geometry of the legs. Ad-

ditionally, the nonbalanced nature of the measuring devices

results in large differences for an arm or a leg between the fit

oftherawdata toaCole modelover thefrequency range. This

is an interesting result and should be investigated further.

The assessment of unilateral lymphedema is performed by

comparing the measured impedance ratios for patients with

lymphedema to the average ratio and the standard spread of

the ratio collected for a normal healthy population. The L-Dex

parameter allows impedance ratios to be compared across

genders and limb dominance on the same scale and is used in

the remainder of the analysis. The agreement between

bioimpedance indices (L-Dex) and interlimb volume differ-

ences, as determined by perometry, for assessment of unilat-

eral arm lymphedema has been reported.12Arm impedance

and volume was measured in 45 women with lymphedema

and 21 women as part of a control group without lymphe-

dema.L-Dexscoreswerehighlycorrelatedwiththedifference

inarmvolumemeasuredbyperometry.Thus,L-Dexprovides

a measurement index that is highly correlated (r¼0.926) with

quantitative measuresofthevolumeincreaseinlimbsizeseen

in lymphedema.

The L-Dex parameter uses the measured limb ratio to give

an indication of the degree of difference of the measured limb

ratio from a normal healthy population limb ratio. L-Dex

values that lieoutside thenormalrange (-10 to 10) orthathave

changedþ10 L-Dex units from baseline may indicate early

signsoflymphedema.Duetotheincreaseinthenaturalspread

of the impedance ratio in a normal population, the sensitivity

of the method is expected to decrease as the measurement

frequency increases and there becomes less clear separation

between a control subject and a lymphedema subject.

The spread of the L-Dex parameter for the control data,

along with the spread of the L-Dex parameter for the sample

lymphedema populations, is shown visually through a box

and whisker plot in Figure 2 for a collection of frequencies for

arms and legs. The L-Dex score is calculatedat each frequency

for each subject as an indicator of lymphedema based on

gender and limb dominance. This allows all subjects to be

plotted on a single graph, despite these differences. The box

and whisker plot displays the sample minimum, the lower

quartile, the median, the upper quartile, and the sample

maximum of the population.

10

1

10

2

10

3

0.032

0.034

0.036

0.038

0.04

0.042

0.044

0.046

0.048

0.05

0.052

Standard Deviation of Z Ratio

Frequency (kHz)

Arm

Leg

FIG. 1.

Standard deviation of impedance ratio for female arms (- -) and legs (-) as a function of current injection frequency.

96 GAW ET AL.

Page 5

Figure 2 clearly shows that there is separation between the

control and test group at R0 for both arms and legs. This

indicates good differentiation between lymphedema subjects

and unaffected subjects for an L-Dex calculated from R0. At

about 20–30kHz, an overlap between the upper quartile of

the normal population and the lower quartile of the affected

arm population becomes evident. This overlap increases as

the measurement frequency increases. In the case of the arms,

the overlap in impedance ratio between the population

groups is larger than for the legs. In both the arm and leg

cases, the overlap between control and test groups becomes

large enough to reduce the sensitivity of assessment above

30kHz.

This indicates that impedance measurements above 30k

kHzarenotsufficientlysensitivetoextracellularfluidchanges

to be used effectively to assess early stages of lymphedema

accurately. This is because the electric current will pass

through a combination of both extra- and intracellular fluids

at these frequencies. The assessment of lymphedema using

BIS has been reported14and has shown that the extracellular

fluid volumes of the arm calculated from R0 are more sensi-

tive in determining the presence of lymphedema than total

body fluids of the arm calculated from the characteristic fre-

quency, Zc (usually around 50–80kHz). The study showed

that the sample populations of the total fluid arm ratio for the

control group and the lymphedema group overlapped.

05 10 2030

Frequency (kHz)

5075 100200500

-20

0

20

40

60

80

100

120

140

160

Arm L-Dex

05 102030

Frequency (kHz)

5075 100200 500

-20

0

20

40

60

80

100

120

140

160

Leg L-Dex

•

•

a

b

FIG. 2.

and clinically evident unilateral arm lymphedema subjects (R) and (b) control (L) and clinically evident unilateral leg

lymphedema subjects (R)

Box and whisker plots demonstrating the spread of L-Dex at a number of increasing frequencies for (a) control (L)

LYMPHEDEMA ASSESSMENT: OPTIMAL IMPEDANCE FREQUENCY 97

Page 6

However, clear discrimination between the two subject

groups is clearly shown for extracellular fluid ratios of the

arms. Therefore, bioimpedance devices that measure the total

fluidofalimbwillnotbesensitivetochangesinlymphedema.

High concordance(rc>0.973

200kHz) has been reported6between impedance ratios cal-

culated at a specific frequency from those calculated at R0for

both arms and leg. However, widening of the 95% confidence

interval and an increased bias are also reported for six in-

creasing frequencies, based on Bland and Altman limits of

agreement. This previously reported method of analysis was

used with the data collected in the current study. Agreement

between the L-Dex calculated at a specific frequency and the

L-Dex calculated at R0was determined using a Bland and

Altman analysis for each of 256 frequencies from 3kHz to

1000kHz. A summary of the mean difference and the 95%

confidence interval as a function of frequency is shown in

Figure 3, separating the arms and legs of the control popula-

tion, the arm lymphedema population, and the leg lymphe-

dema population.

The mean difference between the L-Dex calculated from

R0and the L-Dex calculated for each frequency for both the

arms and the legs of the control population remains constant

for frequenciesbelow

at close to zero for the entire frequency range. The mean

difference for both the arms and legs of the lymphedema

population is shown to increase as frequency increases (0.1

to 12 for arms and ?0.5 to 9 for legs over the frequency range

3kHz–1000kHz). This is because of the presence of fibrotic

tissue expected to be present in an established lymphedema

population such as the one used in this study. At low fre-

quencies, the measurement of impedance comprises pri-

marily extracellular fluid, while impedance measurements

made at high frequencies include interactions from the cells

and thus comprises both tissue and extracellular fluid com-

ponents. Thus the increasing difference between L-Dex at

higher frequencies and the L-Dex calculated from R0is due

to the presence of fibrosis which is not present in a healthy

population. The bias of the mean difference at high fre-

quencies highlights the importance of using low frequency

impedance for the assessment of lymphedema in order to

detect the onset of the disease early and avoid irreversible

damage.

Each population also displays a large widening of the 95%

confidence intervals. This is larger in the lymphedema pop-

ulations. This further demonstrates that the sensitivity of the

method decreases as the current frequency increases.

10

0

10

Frequency (kHz)

1

10

2

10

3

-5

0

5

L-Dex Difference - Control Arm

Mean Difference

95% Confidence Interval

10

0

10

Frequency (kHz)

1

10

2

10

3

-5

0

5

L-Dex Difference - Control Leg

Mean Difference

95% Confidence Interval

10

0

10

Frequency (kHz)

1

10

2

10

3

-20

-10

0

10

20

30

40

L-Dex Difference - Lymphedema Arm

Mean Difference

95% Confidence Interval

10

0

10

Frequency (kHz)

1

10

2

10

3

-20

-10

0

10

20

30

40

L-Dex Difference - Lymphedema Leg

Mean Difference

95% Confidence Interval

ab

cd

FIG. 3.

as a function of frequency for (a) control subject arms, (b) control subject legs, (c) test lymphedema subject arms, and (d) test

lymphedema subject legs.

Summary of Bland and Altman mean difference from L-Dex calculated from R0(- -) and 95% confidence intervals (-)

98GAW ET AL.

Page 7

Conclusions

R0has been shown previously as the best measurement

frequency to assess lymphedema of the arms, as it is more

accurate in measuring the extracellular water content of a

limb.3The present study has shown that the standard devia-

tion of contralateral limb impedance ratios in a normal pop-

ulation increases as the measurement frequency increases. As

such, cut off criteria used to identify the onset of lymphedema

also increases and sensitivity is reduced.

The spread of the control L-Dex and the L-Dex calculated

for both arm and leg test subjects, has also been shown to

increase with frequency. The frequency at which an overlap

occurs between the lower quartile of the test group L-Dex and

the upper quartile of the control group L-Dex begins at about

30kHz.This resultsin a decreased sensitivity for thedetection

of lymphedema above this frequency.

A comparison between the L-Dex calculated at 0kHz

and the L-Dex calculated at any non-zero single fre-

quency shows that the mean difference between the

methods remains constant over the frequency range in a

healthy population and increases in the test lymphedema

populations. This demonstrates the importance of using

low frequency impedance measurements to provide an

accurate assessment of the presence of lymphoedema and

provides further supporting evidence that the current

frequency used in bioimpedance should be below 30kHz

to accurately assess unilateral lymphedema. The 95%

confidence interval was also shown to increase with fre-

quency for each population group. This again shows an

improved sensitivity in assessing lymphedema at low

measurement frequencies.

The many single frequency body composition devices

available on the market traditionally measure the imped-

ance at 50kHz. This study shows that the application of

these devices to the assessment of lymphoedema is less

sensitive to extracellular fluid changes. It has been demon-

strated that changes in the extracellular fluid in unilateral

lymphedema patients (both arm and leg) can only be

monitored accurately and reliably using bioimpedance de-

vices with BIS capabilities or applied current frequencies

below 30kHz.

Acknowledgments

The research studies have been funded by in kind sup-

port from ImpediMed Ltd, Brisbane, Australia, with clinical

measurements of test subjects obtained independently by

Dr. R Box.

Author Disclosure Statement

At the time of data collection, analysis, and reporting, R.

Gaw was also employed at ImpediMed Limited on a part-

time basis. To ensure no conflict of interest existed, clinical

measurements of test subjects were obtained independently

by Dr. R. Box, measurements of control subjects were ob-

tained by R. Gaw and Prof. B. Cornish, and data analysis was

performed by R. Gaw and Prof. B. Cornish. No other com-

peting financial interests exist.

References

1. Hayes S, Janda M, Cornish B, Battistutta D, Newman B.

Lymphedema secondary to breast cancer: How choice of

measure influences diagnosis, prevalence, and identifiable

risk factors. Lymphology 2008;40:18–28.

2. Ridner SH, Dietrich MS, Deng J, Bonner CM, Kidd N.

Bioelectrical impedance for detecting upper limb lymphe-

dema in nonlaboratory settings. Lymphat Res Biol 2009;7:

11–15.

3. Warren AG, Janz BA, Slavin SA, Borud LJ. The use of

bioimpedance analysis to evaluate lymphedema. Ann Plast

Surg 2007;58:541–543.

4. Cornish BH, Thomas BJ, Ward LC. Improved prediction of

extracellular and total body water using impedance loci

generated by multiple frequency bioelectrical impedance

analysis. Phys Med Biol 1993;38: 337–346.

5. Cornish B. Bioimpedance analysis: Scientific background.

Lymphat Res Biol 2006;4:47–50.

6. York SL, Ward LC, Czerniec S, Lee MJ, Refshauge KM,

Kilbreath SL. Single frequency versus bioimpedance spec-

troscopy for the assessment of lymphedema. Breast Cancer

Res Treat 2009;117:177–182.

7. Seo A, Rys M, Konz S. Measuring lower leg swelling: Op-

timum frequency for impedance method. Med Biol Eng

Comput 2001;39:185–189.

8. Oliveira J, Cesar TB. Influence of complex descongestive

physical therapy associated with intake of medium-chain

triclycerides for treating upper-limb lymphedema. Rev Bras

Fisioter 2008;12:31–36.

9. Cornish BH, Jacobs A, Thomas BJ, Ward LC. Optimizing

electrode sites for segmental bioimpedance measurements.

Physiol Meas 1999;20:241–250.

10. Miller AJ, Bruna J, Benninson J. A universally applicable clin-

ical classification of lymphoedema. Angiology 1999;50:189–192

11. Ward LC, Czerniec S, Kilbreath SL. Quantitative bioimpe-

dance spectroscopy for the assessment of lymphoedema.

Breast Cancer Res Treat 2009;117:541–547.

12. Ward LC, Czerniec S, Kilbreath SL. Operational equivalence

of bioimpedance indices and perometry for the assessment of

unilateral arm lymphoedema. Lymphat Res Biol 2009;7:81–85.

13. Cornish BH, Chapman M, Hirst C, Mirolo B, Bunce IH,

Ward LC, Thomas BJ., Early diagnosis of lymphoedema

using multiple frequency bioimpedance. Lymphology 2001;

34:2–11.

14. Cornish BH, Bunce IH, Ward LC, Jones LC, Thomas BJ.

Bioelectrical impedance for monitoring the efficacy of lym-

phoedema treatment programmes. Breast Cancer Res Treat

1996;38:169–176.

Address correspondence to:

Professor Bruce Cornish

Physics Discipline

Faculty of Science and Technology

Queensland University of Technology

2 George St, Brisbane QLD 4001

Brisbane

Australia

E-mail: b.cornish@qut.edu.au

LYMPHEDEMA ASSESSMENT: OPTIMAL IMPEDANCE FREQUENCY99

Page 8

Page 9

This article has been cited by:

1. Stanley G. Rockson . 2011. Maturational Aspects of Postnatal LymphangiogenesisMaturational Aspects of Postnatal

Lymphangiogenesis. Lymphatic Research and Biology 9:2, 75-75. [Citation] [Full Text] [PDF] [PDF Plus]

#### View other sources

#### Hide other sources

- Available from Bruce H Cornish · May 17, 2014
- Available from qut.edu.au