|Home | About | Journals | Submit | Contact Us | Français|
Single-frequency bioelectrical impedance has been used in clinical and research settings to measure extracellular fluid in arms. Its ease of use and low risk of user error suggests this measurement method may have advantages for use in nonlaboratory (community-based) environments when compared to other measurement methods. The purpose of this study was to evaluate the feasibility of using single-frequency bioelectrical impedance to detect upper limb lymphedema in nonlaboratory settings.
Using a standardized protocol, impedance ratios among healthy normal women, breast cancer survivors with lymphedema, and breast cancer survivors without lymphedema were compared with participants seated in an upright position conducive for use outside laboratory settings (community-based environments). Ratios of healthy normal controls and breast cancer survivor groups without lymphedema were very similar, with almost complete overlap in confidence intervals. However, those values were markedly different from the values assessed in the survivor group with lymphedema (p < 0.001).
These findings suggest impedance ratios determined by single-frequency bioelectrical impedance can be used as markers for lymphedema in nonlaboratory settings when a standardized protocol is used.
Approximately 2.4 million breast cancer survivors reside in the United States.1 Despite the development of breast conserving surgical procedures and changes in axillary dissection techniques, current studies suggest that breast cancer survivors remain at risk for development of treatment-related lymphedema, that is, excessive fluid and protein accumulation in the interstitial spaces.2 Recent studies suggest that 6%–40% of breast cancer survivors will develop lymphedema at some point during their lifetime.3–6 These rates include the 6.9%–22% lymphedema occurrence in women undergoing sentinel node biopsies.5,6 Investigation of cancer survivorship issues and the late effects of cancer treatment, such as lymphedema, are emerging as important focal areas for oncology researchers. By necessity, much of this work is being conducted outside traditional laboratory or clinical settings. To date, little work has been published in the use of lymphedema measurement methods in community-based research studies that take place outside the confines of research laboratories or patient care clinics.
Researchers have used multiple methods of evaluating limb volume in laboratories or in clinic settings; these methods include water displacement, circumferential measurement, infrared scanning, and single and multi-frequency bioelectrical impedance.11 Each measurement method has advantages and disadvantages for use in nonlaboratory settings. Water displacement, regarded as a sensitive and accurate “gold standard” for volume measurement in laboratory settings can be used in both clinical and home-based environments.7 However, it is seldom used in these environments because it is cumbersome, messy, and requires the removal of clothing covering the limb. Additionally, the placement of the limb for measurement requires individuals to bend over or to sit with their arms in a cylinder of water. Water displacement cannot be used with patients having open wounds or sores.8
Circumferential measurement of limb volume correlates well with water displacement; however, patients must remove clothing from the limb as they do with water displacement.9 Although portable and inexpensive, circumferential measurement is a time-intensive method and concerns exist about the reliability of the measurements. For example, even in laboratory settings, trained individuals measuring limbs with a tape measure may hold the tape very tightly or loosely around the limb, causing variation in circumferential measurement. Thus, when these measurements are used to calculate limb volume or girth, considerable measurement error occurs. These errors can potentially mask lymphedema progression or falsely imply lymphedema.10
Infrared Perometry, although expensive, is equally or more reliable than circumferential tape measurement methods.8,12 Infrared Perometry provides a quick, hygienic method of determining limb volume. Because the equipment does not touch the skin, measurements can be completed on limbs with sensitive or broken skin. However, it is difficult to obtain accurate measurements in individuals who cannot maintain a stable position during measurement. The size and nonportable nature of the most widely used model requires patients to come into a clinic or research facility for limb measurement.
Portable single and multi-frequency bioelectrical impedance devices have been used for years in research settings in the United States and they recently received Food and Drug Administration (FDA) approval for use in clinic settings in the United States.13–16 These devices can estimate extracellular volume.17 They also provide a limb index ratio (LIR) that represents the impedance difference between affected and unaffected limbs. Australian researchers found the mean LIR to be 1.037 when the dominant arm is the affected arm; and 0.964 when the nondominant arm is the affected arm when using multi-frequency impedance.14 LIR readings of 1.139 or greater for affected dominant arms and 1.066 or greater for affected nondominant arms are thought to be indicators of lymphedema.14,16 Impedance measurements cannot be used for individuals with pacemakers and certain metal implants. When using single-frequency devices to measure limb volume, lightly adhesive electrodes are placed on each hand, each wrist, and on one foot. The placement of electrodes using fixed anatomical landmarks reduces the risk of user error. FDA-approved single-frequency devices that fit in the palm of a hand are available.12 The procedure takes less than 1 minute, it is painless, and clothing does not need to be removed from the limbs. The devices are smaller and less expensive than Perometry.
Single-frequency bioelectrical impedance uses a low voltage electric current to determine extracellular fluid.12,15 The portability of impedance devices poses significant advantages for use in nonlaboratory, community-based environments such as private homes or offices. Measurement does not require individuals to remove any clothing except for foot coverings. Participants do not have to bend over, reducing the risk of falls. If used in a sitting position, participants avoid being asked to assume a more psychologically vulnerable lying-down position. Thus, the purpose of this study was to evaluate the feasibility of using single-frequency, bioelectrical impedance to detect upper limb lymphedema in nonlaboratory settings. Specifically, the primary aim of this study was to compare impedance ratios among breast cancer survivors with and without lymphedema, with that of healthy normal women. All participants were seated in an upright position, a position that could be used outside laboratory settings (community-based environments). A secondary aim was to compare dominant and nondominant arm LIRs to respective normative values established in Australia.
The cross-sectional bioelectrical impedance measurements summarized in this article were collected over the course of three studies. Following Institutional Review Board approval, using multiple methods (e.g., flyers, mass e-mailing, brochures, and word-of-mouth), participants were recruited for these studies from a metropolitan area and the surrounding rural counties in the southeastern United States. Written informed consent was obtained prior to study enrollment. In one study, participants were recruited for comparative research of symptoms and quality of life between breast cancer survivors with and without treatment related lymphedema (n = 149). Findings have been presented elsewhere.18,19 The focus of a second study was the collection of normative impedance values in healthy controls, (n = 60), who were recruited solely for that research. Baseline impedance assessments from a longitudinal intervention study of breast cancer survivors with lymphedema that is currently in progress comprises the third set of values used in this report (n = 24).
Data from all studies were de-identified before analysis. Inclusion in this analysis required that all participants be 18 years of age or older. The participants with a history of breast cancer treatment had (a) a diagnosis of lymphedema in one arm only (lymphedema group) or (b) no diagnosis of lymphedema (nonlymphedema group). Healthy control subjects had no history of lymphedema or breast cancer. Individuals with metal implants and/or pacemakers were excluded, based upon recommendations of the impedance device manufacturer. Pregnant women and women with renal failure or heart failure were also excluded because of patterns of fluid fluctuation associated with these medical conditions. The resulting comparative samples consisted of 60 healthy controls, 98 breast cancer survivors with lymphedema, and 75 breast cancer survivors without lymphedema.
Demographic questionnaire: A self-report questionnaire was used to capture demographic information. Variables included age, race, dominant arm, marital status, annual income, work status, education level, type of insurance, and geographical area of residence.
Height: A Harpenden pocket stadiometer distributed by Seritex Inc. (East Rutherford, New Jersey) was used to measure height. This device provided measurement of height up to 7 feet tall (inches or cm) to within 1 cm.
Bioelectrical impedance: A single-frequency bioelectric impedance device manufactured by Impedimed (Mansfield, Australia) was used to measure extracellular fluid volume.13,16 This instrument provided the ml difference in fluid between the affected and nonaffected limbs and gave a LIR between affected and unaffected limbs.
Participants completed the demographic form and study staff verbally confirmed the information. Height was then measured twice. The average of the two height measurements was then entered into the impedance device, along with data about dominant limb and gender. Because the devices used in this study required information be entered about which arm was the affected arm, the following designations were made. The actual affected (swollen) limb for the participants with lymphedema was entered and a randomly assigned affected arm for all healthy controls was entered for those participants. The ipsilateral arm to the breast cancer site was designated as the affected arm for breast cancer survivors without lymphedema.
During impedance measurements, all participants were seated at nonmetal desks with their feet resting on the floor. A hard foam cushion was placed under the arms slightly below shoulder level. Participants rested their arms with palms facing down on the cushion and were instructed to relax as much as possible during the measurement. The skin in areas where electrodes were to be applied was wiped with an alcohol pad and allowed to dry. Then lightly adhesive resting electrodes were placed on each hand at the dorsal surface of the wrist between the process of the radial and the ulnar bones and on the dorsal surface of the hand, 1 cm proximal from the peak of the knuckle of the middle finger. A foot electrode was placed midway between the two bony processes on the ankle in the front of the foot. Color-coded alligator clips were placed on targeted electrodes. Measurement of the actual or designated affected arm was made first; the alligator clips were then switched and measurement of the other limb was taken. The data were then recorded. The average time to complete the impedance measurement was less than 5 min, including participant preparation. LIRs representing the difference in volume or impedance between the arms of the participants were recorded (the cancer treatment affected and nonaffected arm in survivors and between the designated affected and nonaffected arm in healthy controls).
Statistical data analyses were conducted using SPSS (version 15.0, SPSS Inc., Chicago, IL) and Stata (Release 9, Stat-aCorp LP, College Station, TX). Descriptive statistics were used to summarize demographic characteristics and LIRs. Some of the LIR distributions were not normally distributed. Specifically, some LIRs for the breast cancer survivors with lymphedema were extreme, generating positively skewed distributions. Thus, median and interquartile range (IQR) statistics are included in the descriptive summaries and Kruskall–Wallis tests were used to test for overall differences among the arm samples. Post-hoc analyses of statistically significant differences were conducted using Mann–Whitney pairwise comparisons with a Bonferroni-adjusted alpha value. To diminish the possible effect of extreme LIR values on population estimates using these samples, bias-corrected bootstrapped 99% confidence intervals were calculated and are displayed in Figure 1. Bias-corrected, standard errors are presented rather than standard deviations, which assume a normal distribution.
The sample (n = 233) consisted of females ages 30 through 94 years old. Sample characteristics are presented in Table 1 and reflect that 47 individuals opted not to report their annual income and that one individual did not wish to be classified by race.
Within the sample of 60 healthy controls, 32 (53%) of LIR assessments were done using the dominant arm as the numerator, and 28 (47%) were done using the nondominant arm as the numerator for the LIR. The affected side of 45 (46%) survivors with lymphedema and 43 (57%) survivors without lymphedema was the dominant arm side and thus LIR numerators were for the dominant arm. Finally, the non-dominant arm was the affected side for 53 (54%) breast cancer survivors with lymphedema and 32 (43%) breast cancer survivors without lymphedema, and LIRs were established using that side as the numerator. Detailed descriptive summaries of the LIRs are displayed in Table 2. The LIRs for the healthy controls and breast cancer survivors without lymphedema were very similar in both average values (approximately 1.02 for dominant and 0.980 for nondominant arms) and variability of the LIRs (SEMs of 0.005–0.007). In addition, these are very similar to the Australian normative values of 1.037 for dominant arm assessments and 0.964 for non-dominant arms. However, LIRs for the breast cancer survivors with lymphedema were markedly and statistically significantly higher than the respective nonlymphedema arm assessments in this research (p < 0.001) and from the Australian normative values. The bias-corrected bootstrapped 99% confidence intervals for each mean LIR are displayed in Figure 1, along with a line designating the respective normative values established in Australia. As depicted, LIR population estimates for healthy controls and breast cancer survivors without lymphedema are very similar with almost complete overlap in confidence intervals. In contrast, the population estimates of LIRs for breast cancer survivors with lymphedema are considerably higher and, despite those values being much more varied and thus the 99% confidence intervals considerably less precise, no overlap with the nonlymphedema intervals (p < 0.01) was demonstrated. These findings suggest that bioelectrical impedance can be used outside laboratory settings to evaluate arm lymphedema and that LIRs can be used with confidence as markers for lymphedema.
Bioelectrical impedance measurement of arms has reduced user error due to use of specified anatomical landmarks for electrode placement and standardized impedance formulas contained within its programmed software. The protocol used in this study was easy for participants to follow. It took less than 5 min and was well-tolerated by individuals of all ages and size. Therefore, use of bioelectrical impedance to measure limbs in nonlaboratory based settings may be appealing to researchers. It is also interesting to note that several participants in this study expressed interest in using bioelectrical impedance at home to measure their arms regularly as part of their self-care. They thought the device appeared to be simple to use and that electrodes and leads could easily be applied by themselves or a family member.
The procedure implemented in this cross-sectional study resulted in bioelectrical impedance values that clearly distinguished between individuals with and without lymphedema and LIR means similar to those found in previous studies. These findings suggest that in addition to the ease of use and rapidity of measurement, bioelectrical impedance can be used outside laboratory settings to evaluate arm lymphedema. Findings also suggest that LIRs can be used with confidence as markers for lymphedema.
Future impedance research is indicated. Specifically, given the cross-sectional nature of this study, additional studies using this technology to examine limb volume fluctuations over time in individuals with and without lymphedema are indicated. Additionally, given the interest expressed by participants of integrating bioelectrical impedance measurement of their arms into their home-based self-care regimens, feasibility studies exploring such use are warranted.
This research was funded by National Research Service Award 1F31NR07854-02, a Vanderbilt University School of Nursing Oncology Postdoctoral Fellowship, and American Cancer Society Grant MRSG-07-012-01-CPPB.
Drs. Ridner and Dietrich and Ms Deng, Ms Bonner, and Ms Kidd have no conflicts of interest or financial ties to disclose.