- Email: [email protected]
Detection of segmental internal fat by bioelectrical impedance analysis in a biological phantom
BASIC NUTRITIONAL INVESTIGATION
Detection of Segmental Internal Fat by Bioelectrical Impedance Analysis in a Biological Phantom Qing He, MS, Jack Wang, MS, Ellen S. Engelson, EdD, and Donald P. Kotler, MD From the Division of Gastroenterology and the Body Composition Unit, Department of Medicine, St. Luke’s–Roosevelt Hospital Center, Columbia University College of Physicians & Surgeons, New York, New York, USA OBJECTIVE: Quantification of internal adipose tissue such as visceral adipose tissue currently relies on expensive, cross-sectional imaging modalities. The purpose of this study was to test the hypothesis that surface impedance, determined by bioimpedance analysis, might be used to predict regional internal fat content change in a phantom model. METHODS: Fresh hollowed-out cucumbers were used as cylindrical biological phantoms to test this hypothesis. After removal of the seeds, the cucumbers were filled with normal saline, mixture of saline and corn oil, or porcine adipose tissue bathed in saline. Surface resistance and reactance were measured with a bioimpedance analyzer accurate to 0.1 ⍀ (Quantum 10X, RJL Systems), and impedance was calculated. A linear regression model was used to interpret the association between composition and impedance. RESULTS: Surface impedance varied linearly with changes in the relative internal corn oil portions (r– ⫽ 0.98). A similar relation was noted with porcine adipose tissue bathed in saline (r2 ⫽ 0.95) regardless of the specific position of adipose tissue within the cucumber. CONCLUSION: Surface impedance measured by bioimpedance analysis can detect variations in fat content in the interior of a cylindrical phantom. Nutrition 2003;19:541–544. ©Elsevier Inc. 2003 KEY WORDS: bioelectrical impedance analysis, regional fat, phantom, body composition
INTRODUCTION Various body fat depots have independent metabolic implications and associated health risks. In particular, increased visceral adipose tissue (VAT) is associated with insulin resistance, hyperlipidemia, and increased risk of cardiovascular disease.1–5 For this reason, there is intense interest in quantitating VAT and its change in relation to cardiovascular risk factors. However, this compartment is difficult to measure accurately due to its deep anatomic location and irregular distribution inside the abdominal cavity. Several methods have been used. Anthropometric variables such as waist circumference, waist-to-hip ratio, and sagittal diameter are simple and easy to determine but have substantial predictive errors and are neither sensitive nor specific for estimating VAT or its change.6,7 A more technologically sophisticated method, dualenergy X-ray absorptiometry, estimates total body and regional fat without differentiating abdominal subcutaneous fat from visceral fat; its advantage over anthropometry in measuring VAT is controversial.8,9 Imaging modalities, such as magnetic resonance imaging and computed tomography, have been used in various sex, age, and weight groups to estimate regional fat contents.10 –13 They are regarded as the current “gold standard” for determination of visceral fat. However, such instruments are expensive and of
This study was funded fully by Serono Laboratories, Inc. Correspondence to: Donald P. Kotler, MD, S&R Building 1301, 1111 Amsterdam Avenue, New York, NY 10025, USA. E-mail: dpkotler@ aol.com Nutrition 19:541–544, 2003 ©Elsevier Inc., 2003. Printed in the United States. All rights reserved.
limited accessibility, and analysis of images is time consuming. Bioelectrical impedance analysis (BIA) has been used to measure total body fat, fat-free mass (FFM), and regional FFM.14 –18 However, the potential of BIA to estimate segmental internal fat such as VAT has less well been validated. The BIA measures, resistance and reactance, are based on the conduction of a low-voltage current by soluble ions. Conduction of electrical current is primarily through FFM. Consequently, biological conductor could be divided into current compatible FFM and residual fat compartment. Conductivity is related to the amount of FFM. The human torso can be regarded as a two-compartment model consisting of FFM and fat in parallel. As in other models of regional body composition,14,15 fat content equals the difference between physical volume and electrical conductor volume if a simple relation is established between the latter and FFM. The model is valid if the BIA current passes through the center of the torso. Cross-sectional studies have shown that impedance index (calculated as length2/impedance) in the trunk area is directly related to FFM.17,18 Extrapolation of this finding to changes in body composition requires establishment of a similar relation longitudinally. The purpose of this study was to explore the ability of BIA to reflect composition change in the core of a biological cylinder. To mimic the human peritoneal cavity, we used a hollowed-out cucumber filled with various proportions of mixtures of corn oil and saline or pieces of porcine adipose tissue bathed in saline. By using surface electrodes, we determined the effect on surface impedance of differing internal fat and saline content. We also evaluated whether the length of the surface electrodes affects measured impedance and the ability to detect changes in internal content. 0899-9007/03/$30.00 doi:10.1016/S0899-9007(03)00038-8
542
He et al.
Nutrition Volume 19, Number 6, 2003 500-⍀ standard resistor provided by the BIA manufacturer. Each record was the mean value of three readings. We validated the Quantum 10X with the widely used BIA 101A (RJL Systems) by paired t test. Because there was no significant difference between the two systems, we report only Quantum 10X data from the following experiments. Preexperiment Procedures All cucumbers were treated with alcohol to remove surface wax, rinsed with water, and dried with clean paper before electrode placement. The circumferences at the four electrode locations were determined by a tape measure, and the mean of the four circumferences was regarded as the circumference of the cucumber. All cucumbers were positioned upright, as shown in Figure 1, and new cucumbers were used for each experiment. Experiment 1
FIG. 1. Schematic drawing of a cucumber phantom. BIA, bioelectrical impedance analysis.
The aim of experiment 1 was to validate the cucumber as a biological model of electrolyte-containing compartments. The resistance and reactance of cucumbers were measured, and then each cucumber was covered with a thick piece of moistened tissue and steamed over mildly boiling water for 10 min. Steaming served as a means to destroy the cell structure and homogenize the cucumber. Cucumbers that had intact skin after steaming were selected and cooled at room temperature for 3 h. Resistance and reactance were measured again. Experiment 2
MATERIALS AND METHODS Materials We conducted a series of experiments on selected cucumbers 18 to 25 cm long, with circumferences of 18 to 24 cm. The middle segments of the cucumbers were nearly cylindrical in shape. The outer skins of the cucumbers were intact, and surfaces were generally smooth for the sake of good contact with the electrodes. Each cucumber was dedicated to one experiment. All experiments in the series were replicated on a second cucumber. We used Mazola pure corn oil (Bestfood Canadian Inc., Montreal, Canada). Porcine adipose tissue was purchased fresh at a meat market. Chemical analysis of adipose samples, performed at the Obesity Research Center of St. Luke’s Hospital, indicated that the lean component was negligible (⬍3%). Saline was a laboratory-prepared sodium chloride solution at a concentration of 0.93%. Electrodes were cut from a roll of aluminum tape with adhesive salt gel underneath. The width of electrodes was set at 0.5 in., and length varied. A set of electrodes consisted of four equal-length conductors. Two acted as source electrodes and the other two acted as detecting electrodes. Electrodes were applied horizontally and aligned vertically (Fig. 1). The distance between the source electrode and the detecting electrode at the same end was 1.5 cm. The distance between source electrodes was determined by the length of the cucumber, so that the detection range covered the entire portion of the cucumber cavity that held the different mixtures. Instrument We used a Quantum 10X bioimpedance analyzer (RJL Systems, Clinton Township, MI, USA). This is a new version of tetrapolar BIA machine that is sensitive to 0.1 ⍀. An 800-A alternating current at 50 kHz was administered at the source electrodes, and resistance and reactance were measured at the detecting electrodes. Impedance was calculated as the square root of (resistance2 ⫹ reactance2). BIA readings were preceded by calibration with a
Experiment 2 tested whether BIA-generated electrical current can pass through the center of a cucumber by investigating the influence of its internal content on BIA readings. Each cucumber was prepared by opening a deep cylindrical hole with a diameter of approximately 2.5 cm at the top end, and all seeds were removed. Subsequently, holes were filled sequentially with saline or a mixture of saline and oil. Resistance and reactance of the middle segment were measured. Experiment 3 With the result of experiment 2 available, experiment 3 explored the quantitative relation between corn oil content in a cucumber center and BIA readings. Cucumbers were treated as in experiment 2, and total central volume was determined. We made serial measurements of resistance and reactance while changing the corn oil proportion of the central mixture of saline and oil. Mixtures of the same total volume were poured serially into the center space. Resistance and reactance readings were taken by using electrode lengths of 50% of the cucumber circumference. The same procedure was repeated on a different cucumber with electrode lengths of 25% of the cucumber circumference. Experiment 4 Experiment 4 tested the effect of internal adipose tissue volume on BIA readings. Three porcine adipose tissue samples of different weights were consecutively put into a cucumber chamber prepared as in experiment 2, the rest of chamber was filled with saline to the same level above the electrodes, and resistance and reactance readings were taken. Data Analysis A simple linear regression model was used to analyze the relation between impedance and oil content. Student’s t test was used to
Nutrition Volume 19, Number 6, 2003
Internal BIA Fat in a Phantom
543
TABLE I. EFFECT OF VARIATIONS IN CUCUMBER CONTENT ON IMPEDANCE*
Conditions Baseline (whole cucumber) Seeds removed Filled with saline only Filled with equal amounts of saline and oil Filled with corn oil only
Resistance (⍀)
Reactance (⍀)
Impedance (⍀)†
183.9 209.1 51.1 149.3
167.9 155.4 11.3 78.6
249.0 265.0 52.3 168.7
192.3
116.0
224.0
* Data are from the Quantum 10X bioelectrical impedance analyzer. Electrodes were placed around the circumference of the cucumber at half its length. † Resistance2 ⫹ reactance2.
determine the probability value of  of regression equations. Coefficients of determination (r2) were calculated. P ⬍ 0.05 was used to determine significance of all results.
FIG. 3. Relation between corn oil proportion and surface impedance, with an electrode length of 25% of circumference (total internal volume was 130 mL).
tissue in a saline-filled cucumber (Fig. 4), regardless of its position inside the cucumber (data not shown).
DISCUSSION RESULTS Steaming a fresh cucumber reduced its impedance from 706.2 ⫾ 0.09 (mean ⫾ standard deviation) ⍀ to 108.7 ⫾ 0.04 ⍀, reflecting a 77% fall in resistance and a 98% drop in reactance. There was a significant difference in impedance before and after steaming (P ⱕ 0.01). Removal of the seeds from the cucumber increased impedance by about 6% in experiment 2. Filling the central space with highly conductive saline significantly reduced impedance by about 80%. Subsequent replacement of saline with corn oil significantly raised impedance (Table I), suggesting that internal content did influence impedance and that electrical current did go through the center of the cucumber. Replacing some saline with corn oil resulted in a proportional increase of surface impedance of whole cucumbers. The results were similar at electrode lengths of 50% and 25% of cucumber circumference (Figs. 2 and 3). The data set fit well with the linear regression model (r2 ⫽ 0.98 and 0.97, respectively). Occupation of the central cucumber chamber with porcine adipose tissues in a similar experiment showed a similar relation. Impedance increased proportionally to the mass of porcine adipose
FIG. 2. Relation between corn oil proportion of mixture and surface impedance, with an electrode length of 50% of cucumber circumference (total internal volume was 100 mL).
The major finding of this study is that a 50-Hz bioimpedance analyzer can detect a composition change within a biological cylinder, indicating that the electrical current passes through the entire cucumber, not only the peripheral “FFM” but also the central content, and recruits the corresponding conducting material. This is a requirement for extrapolation of our cucumber model to the human torso model. There are reports of segmental bioimpedance analysis to detect fat-free tissue.14,15 Our study explores relations between changes in the segmental ratio of FFM to volume and impedance. Both applications of regional BIA rely on the same principle. Lean tissue can be considered an electrical volume that can be compared to physical volume, with the difference related to non-conductive tissue. In a simple two-compartment model, the higher the difference in resistance between the two compartments, the greater the relative amount of current that will pass through the less resistive conductor. Because corn oil is non-conductive, virtually all of the electrical current presumably went through the conducting media of saline and cucumber mass. The distinction between lean and non-lean mass by BIA would be closer to that defined by anatomy. In the case of the human torso, for which the cucumber was used as a biological validation, the non-conductive material is comprised mainly of adipose tissue.
FIG. 4. Relation between porcine adipose tissue and surface impedance, with an electrode length of 50% of cucumber circumference.
544
He et al.
Nutrition Volume 19, Number 6, 2003
We proved our cucumber model is a valid biological model through experiment 1. Our cucumbers responded to steaming in much the same way as an earlier potato biological model validated by Lukaski using similar measurements.19 The steam-induced resistance drop in our cucumber model was 77% and reactance drop was 98%, which was close to the 76% and 100% drop documented in cooked versus raw potatoes. The results suggested that steaming enhances conductivity and abolishes reactance by destroying the cell compartments. To our knowledge, our further experiments are novel and have not been conducted previously. We found that surface impedance changed proportionally with the quantity of oil inside the cucumber. This relation was reproduced in cucumbers of different sizes with variable electrode length to circumference ratios. Therefore, the relation between impedance and internal composition is an inherent characteristic of this model, even though electrode length did affect the absolute BIA readings (data not shown). Because variation in corn oil portion in our model is equal to the variation in FFM relative to the total physical volume, a linear relation between the ratio of FFM to volume and impedance can also be interpreted as a linear relation between electrical volume and impedance regardless of total physical volume. Because fat in humans exists largely in solid form within adipose tissue, we also evaluated our model by testing solid porcine adipose tissue by methods similar to those to test liquid corn oil. We continued to see a linear relation between surface impedance and internal fat content when the fat was in a solid form regardless of its position within the BIA measurement range. These studies in cucumbers did not establish that the changes would be similar in humans. Our goal of generating a method using BIA for the clinical prediction of visceral fat content in future studies faces many potential roadblocks. Although these results showed that surface impedance reflects changes in conducting volume (electrical volume) within a segment the size of a cucumber, it remains uncertain whether the larger volume of the human torso would behave similarly. If so, it may be possible to determine non-conductive tissue (fat) volume inside an anatomic segment if one can determine the physical volume with a simple method such as anthropometry. Determining physical volume may be possible if the torso is divided into multiple cylindrical cones. This method of truncating abdominal segments is used in the current magnetic resonance imaging protocol for the whole body and was previously validated.20 Subtracting conductive volume from total physical volume would provide residual volume, which in the human abdomen is composed mainly of adipose tissue and bone. We do not expect that any method using BIA alone will be capable of further separating adipose tissue into subcutaneous and visceral compartments. However, we recently found21 a strong association between subcutaneous adipose tissue on magnetic resonance imaging and an anthropometrics model proposed by Wang et al.22 This suggests that a combination of measurements, including multiple abdominal circumferences, spacing gaps, and skinfold thicknesses, can be used to derive reliable predictions of total abdominal volume and subcutaneous adipose tissue volume. Therefore, it may be possible to estimate VAT size through subtraction. Our model supports the hypothesis that measurement of surface impedance in combination with variables of volume can be used to predict segmental internal fat content should a standard equation be derived to predict electrical volume. However, extrapolation of our simple model to measurement of the human visceral fat compartment requires further investigation, because other variables, such as irregularity of abdominal shape and separation of “subcutaneous” and “visceral” adipose tissue compartments, may affect
the result. Nevertheless, the current report is a promising first step toward the goal of using surface impedance by BIA to determine visceral fat content and its changes over time.
REFERENCES 1. Brochu M, Starling RD, Tchernof A, et al. Visceral adipose tissue is an independent correlate of glucose disposal in older obese postmenopausal women. J Clin Endocrinol Metab 2000;852:2378 2. Larsson B, Svardsudd K, Welin L, et al. Abdominal adipose tissue distribution, obesity, and risk of cardiovascular disease and death: 13 year follow up of participants in the study of men born in 1913. Br Med J 1984;288:1401 3. Nakamura T, Tokunaga K, Shimomura I, et al. Contribution of visceral fat accumulation to the development of coronary artery disease in non-obese men. Atherosclerosis 1994;107:239 4. Pascot A, Despres JP, Lemieux I, et al. Deterioration of the metabolic risk profile in women. Respective contributions of impaired glucose tolerance and visceral fat accumulation. Diabetes Care 2001;24:902 5. Pascot A, Despres JP, Lemieux I, et al. Contribution of visceral obesity to the deterioration of the metabolic risk profile in men with impaired glucose tolerance. Diabetologia 2000;43:1126 6. Thomas EL, Brynes AE, McCarthy J, et al. Preferential loss of visceral fat following aerobic exercise, measured by magnetic resonance imaging. Lipids 2000;35:769 7. Ross R, Shaw KD, Rissanen J, et al. Sex differences in lean and adipose tissue distribution by magnetic resonance imaging: anthropometric relationships. Am J Clin Nutr 1994;59:1277 8. Clasey JL, Bouchard C, Teates CD, et al. The use of anthropometric and dual-energy X-ray absorptiometry (DXA) measures to estimate total abdominal and abdominal visceral fat in men and women. Obes Res 1999;7:256 9. Kamel EG, McNeill G, Van Wijk MC. Usefulness of anthropometry and DXA in predicting intra-abdominal fat in obese men and women. Obes Res 2000;8:36 10. Ross R, Rissanen J, Pedwell H, Clifford J, Shragge P. Influence of diet and exercise on skeletal muscle and visceral adipose tissue in men. J Appl Physiol 1996;81:2445 11. Fowler PA, Fulle MF, Glasbey CA, et al. Total and subcutaneous adipose tissue in women: the measurement of distribution and accurate prediction of quantity by using magnetic resonance imaging. Am J Clin Nutr 1995;54:18 12. Baumgartner RN, Rhyne RL, Troup C, Wayne S, Garry PJ. Appendicular skeletal muscle areas assessed by magnetic resonance imaging in older persons. J Gerontol 1992;47:M67 13. Engelson ES, Kotler DP, Tan Y, et al. Fat distribution in HIV-infected patients reporting truncal enlargement quantified by whole-body magnetic resonance imaging. Am J Clin Nutr 1999;69:1162 14. Heymsfield SB, Gallagher D, Grammes J, et al. Upper extremity skeletal muscle mass: potential of measurement with single frequency bioimpedance analysis. Appl Radiat Isot 19989;49:473 15. Lukaski HC. Assessing regional muscle mass with segmental measurements of bioelectrical impedance in obese women during weight loss. Ann NY Acad Sci 2000;904:154 16. Kotler DP, Burastero S, Wang J, Pierson RN, Jr. Prediction of body cell mass, fat-free mass, and total body water with bioelectrical impedance analysis: effects of race, sex, and disease. Am J Clin Nut 1994;64:489S 17. Bracco D, Thiebaud D, Chiolero RL, et al. Segmental body composition assessed by bioelectrical impedance analysis and DEXA in humans. J Appl Physiol 1996;81:2580 18. Fuller NJ, Fewtrell MS, Dewit O, Elia M, Wells JC. Segmental bioelectrical impedance analysis in children aged 8 –12 y: 2. The assessment of regional body composition and muscle mass. Int J Obes Relat Metab Disord 2002;26:692 19. Lukaski HC. Biological indexes considered in the derivation of the bioelectrical impedance analysis. Am J Clin Nutr 1996;64:397S 20. Abate N, Burns D, Peshock RM, Garg A, Grundy SM. Estimation of adipose tissue mass by magnetic resonance imaging: validation against dissection in human cadavers. J Lipid Res 1994;35:1490 21. He Q, Wang J, Engelson ES, Kotler DP. Ability of anthropometrics model to track changes in subcutaneous adipose tissue area (abstract). FASEB J 2002;16: A1024 22. Wang J, Thornton JC, Russell M, et al. Asians have lower body mass index (BMI) but higher percent body fat than do whites: comparisons of anthropometric measurements. Am J Clin Nutr 1994;60:23
(For an additional perspective, see Editorial Opinions.)
Detection of Segmental Internal Fat by Bioelectrical Impedance Analysis in a Biological Phantom Qing He, MS, Jack Wang, MS, Ellen S. Engelson, EdD, and Donald P. Kotler, MD From the Division of Gastroenterology and the Body Composition Unit, Department of Medicine, St. Luke’s–Roosevelt Hospital Center, Columbia University College of Physicians & Surgeons, New York, New York, USA OBJECTIVE: Quantification of internal adipose tissue such as visceral adipose tissue currently relies on expensive, cross-sectional imaging modalities. The purpose of this study was to test the hypothesis that surface impedance, determined by bioimpedance analysis, might be used to predict regional internal fat content change in a phantom model. METHODS: Fresh hollowed-out cucumbers were used as cylindrical biological phantoms to test this hypothesis. After removal of the seeds, the cucumbers were filled with normal saline, mixture of saline and corn oil, or porcine adipose tissue bathed in saline. Surface resistance and reactance were measured with a bioimpedance analyzer accurate to 0.1 ⍀ (Quantum 10X, RJL Systems), and impedance was calculated. A linear regression model was used to interpret the association between composition and impedance. RESULTS: Surface impedance varied linearly with changes in the relative internal corn oil portions (r– ⫽ 0.98). A similar relation was noted with porcine adipose tissue bathed in saline (r2 ⫽ 0.95) regardless of the specific position of adipose tissue within the cucumber. CONCLUSION: Surface impedance measured by bioimpedance analysis can detect variations in fat content in the interior of a cylindrical phantom. Nutrition 2003;19:541–544. ©Elsevier Inc. 2003 KEY WORDS: bioelectrical impedance analysis, regional fat, phantom, body composition
INTRODUCTION Various body fat depots have independent metabolic implications and associated health risks. In particular, increased visceral adipose tissue (VAT) is associated with insulin resistance, hyperlipidemia, and increased risk of cardiovascular disease.1–5 For this reason, there is intense interest in quantitating VAT and its change in relation to cardiovascular risk factors. However, this compartment is difficult to measure accurately due to its deep anatomic location and irregular distribution inside the abdominal cavity. Several methods have been used. Anthropometric variables such as waist circumference, waist-to-hip ratio, and sagittal diameter are simple and easy to determine but have substantial predictive errors and are neither sensitive nor specific for estimating VAT or its change.6,7 A more technologically sophisticated method, dualenergy X-ray absorptiometry, estimates total body and regional fat without differentiating abdominal subcutaneous fat from visceral fat; its advantage over anthropometry in measuring VAT is controversial.8,9 Imaging modalities, such as magnetic resonance imaging and computed tomography, have been used in various sex, age, and weight groups to estimate regional fat contents.10 –13 They are regarded as the current “gold standard” for determination of visceral fat. However, such instruments are expensive and of
This study was funded fully by Serono Laboratories, Inc. Correspondence to: Donald P. Kotler, MD, S&R Building 1301, 1111 Amsterdam Avenue, New York, NY 10025, USA. E-mail: dpkotler@ aol.com Nutrition 19:541–544, 2003 ©Elsevier Inc., 2003. Printed in the United States. All rights reserved.
limited accessibility, and analysis of images is time consuming. Bioelectrical impedance analysis (BIA) has been used to measure total body fat, fat-free mass (FFM), and regional FFM.14 –18 However, the potential of BIA to estimate segmental internal fat such as VAT has less well been validated. The BIA measures, resistance and reactance, are based on the conduction of a low-voltage current by soluble ions. Conduction of electrical current is primarily through FFM. Consequently, biological conductor could be divided into current compatible FFM and residual fat compartment. Conductivity is related to the amount of FFM. The human torso can be regarded as a two-compartment model consisting of FFM and fat in parallel. As in other models of regional body composition,14,15 fat content equals the difference between physical volume and electrical conductor volume if a simple relation is established between the latter and FFM. The model is valid if the BIA current passes through the center of the torso. Cross-sectional studies have shown that impedance index (calculated as length2/impedance) in the trunk area is directly related to FFM.17,18 Extrapolation of this finding to changes in body composition requires establishment of a similar relation longitudinally. The purpose of this study was to explore the ability of BIA to reflect composition change in the core of a biological cylinder. To mimic the human peritoneal cavity, we used a hollowed-out cucumber filled with various proportions of mixtures of corn oil and saline or pieces of porcine adipose tissue bathed in saline. By using surface electrodes, we determined the effect on surface impedance of differing internal fat and saline content. We also evaluated whether the length of the surface electrodes affects measured impedance and the ability to detect changes in internal content. 0899-9007/03/$30.00 doi:10.1016/S0899-9007(03)00038-8
542
He et al.
Nutrition Volume 19, Number 6, 2003 500-⍀ standard resistor provided by the BIA manufacturer. Each record was the mean value of three readings. We validated the Quantum 10X with the widely used BIA 101A (RJL Systems) by paired t test. Because there was no significant difference between the two systems, we report only Quantum 10X data from the following experiments. Preexperiment Procedures All cucumbers were treated with alcohol to remove surface wax, rinsed with water, and dried with clean paper before electrode placement. The circumferences at the four electrode locations were determined by a tape measure, and the mean of the four circumferences was regarded as the circumference of the cucumber. All cucumbers were positioned upright, as shown in Figure 1, and new cucumbers were used for each experiment. Experiment 1
FIG. 1. Schematic drawing of a cucumber phantom. BIA, bioelectrical impedance analysis.
The aim of experiment 1 was to validate the cucumber as a biological model of electrolyte-containing compartments. The resistance and reactance of cucumbers were measured, and then each cucumber was covered with a thick piece of moistened tissue and steamed over mildly boiling water for 10 min. Steaming served as a means to destroy the cell structure and homogenize the cucumber. Cucumbers that had intact skin after steaming were selected and cooled at room temperature for 3 h. Resistance and reactance were measured again. Experiment 2
MATERIALS AND METHODS Materials We conducted a series of experiments on selected cucumbers 18 to 25 cm long, with circumferences of 18 to 24 cm. The middle segments of the cucumbers were nearly cylindrical in shape. The outer skins of the cucumbers were intact, and surfaces were generally smooth for the sake of good contact with the electrodes. Each cucumber was dedicated to one experiment. All experiments in the series were replicated on a second cucumber. We used Mazola pure corn oil (Bestfood Canadian Inc., Montreal, Canada). Porcine adipose tissue was purchased fresh at a meat market. Chemical analysis of adipose samples, performed at the Obesity Research Center of St. Luke’s Hospital, indicated that the lean component was negligible (⬍3%). Saline was a laboratory-prepared sodium chloride solution at a concentration of 0.93%. Electrodes were cut from a roll of aluminum tape with adhesive salt gel underneath. The width of electrodes was set at 0.5 in., and length varied. A set of electrodes consisted of four equal-length conductors. Two acted as source electrodes and the other two acted as detecting electrodes. Electrodes were applied horizontally and aligned vertically (Fig. 1). The distance between the source electrode and the detecting electrode at the same end was 1.5 cm. The distance between source electrodes was determined by the length of the cucumber, so that the detection range covered the entire portion of the cucumber cavity that held the different mixtures. Instrument We used a Quantum 10X bioimpedance analyzer (RJL Systems, Clinton Township, MI, USA). This is a new version of tetrapolar BIA machine that is sensitive to 0.1 ⍀. An 800-A alternating current at 50 kHz was administered at the source electrodes, and resistance and reactance were measured at the detecting electrodes. Impedance was calculated as the square root of (resistance2 ⫹ reactance2). BIA readings were preceded by calibration with a
Experiment 2 tested whether BIA-generated electrical current can pass through the center of a cucumber by investigating the influence of its internal content on BIA readings. Each cucumber was prepared by opening a deep cylindrical hole with a diameter of approximately 2.5 cm at the top end, and all seeds were removed. Subsequently, holes were filled sequentially with saline or a mixture of saline and oil. Resistance and reactance of the middle segment were measured. Experiment 3 With the result of experiment 2 available, experiment 3 explored the quantitative relation between corn oil content in a cucumber center and BIA readings. Cucumbers were treated as in experiment 2, and total central volume was determined. We made serial measurements of resistance and reactance while changing the corn oil proportion of the central mixture of saline and oil. Mixtures of the same total volume were poured serially into the center space. Resistance and reactance readings were taken by using electrode lengths of 50% of the cucumber circumference. The same procedure was repeated on a different cucumber with electrode lengths of 25% of the cucumber circumference. Experiment 4 Experiment 4 tested the effect of internal adipose tissue volume on BIA readings. Three porcine adipose tissue samples of different weights were consecutively put into a cucumber chamber prepared as in experiment 2, the rest of chamber was filled with saline to the same level above the electrodes, and resistance and reactance readings were taken. Data Analysis A simple linear regression model was used to analyze the relation between impedance and oil content. Student’s t test was used to
Nutrition Volume 19, Number 6, 2003
Internal BIA Fat in a Phantom
543
TABLE I. EFFECT OF VARIATIONS IN CUCUMBER CONTENT ON IMPEDANCE*
Conditions Baseline (whole cucumber) Seeds removed Filled with saline only Filled with equal amounts of saline and oil Filled with corn oil only
Resistance (⍀)
Reactance (⍀)
Impedance (⍀)†
183.9 209.1 51.1 149.3
167.9 155.4 11.3 78.6
249.0 265.0 52.3 168.7
192.3
116.0
224.0
* Data are from the Quantum 10X bioelectrical impedance analyzer. Electrodes were placed around the circumference of the cucumber at half its length. † Resistance2 ⫹ reactance2.
determine the probability value of  of regression equations. Coefficients of determination (r2) were calculated. P ⬍ 0.05 was used to determine significance of all results.
FIG. 3. Relation between corn oil proportion and surface impedance, with an electrode length of 25% of circumference (total internal volume was 130 mL).
tissue in a saline-filled cucumber (Fig. 4), regardless of its position inside the cucumber (data not shown).
DISCUSSION RESULTS Steaming a fresh cucumber reduced its impedance from 706.2 ⫾ 0.09 (mean ⫾ standard deviation) ⍀ to 108.7 ⫾ 0.04 ⍀, reflecting a 77% fall in resistance and a 98% drop in reactance. There was a significant difference in impedance before and after steaming (P ⱕ 0.01). Removal of the seeds from the cucumber increased impedance by about 6% in experiment 2. Filling the central space with highly conductive saline significantly reduced impedance by about 80%. Subsequent replacement of saline with corn oil significantly raised impedance (Table I), suggesting that internal content did influence impedance and that electrical current did go through the center of the cucumber. Replacing some saline with corn oil resulted in a proportional increase of surface impedance of whole cucumbers. The results were similar at electrode lengths of 50% and 25% of cucumber circumference (Figs. 2 and 3). The data set fit well with the linear regression model (r2 ⫽ 0.98 and 0.97, respectively). Occupation of the central cucumber chamber with porcine adipose tissues in a similar experiment showed a similar relation. Impedance increased proportionally to the mass of porcine adipose
FIG. 2. Relation between corn oil proportion of mixture and surface impedance, with an electrode length of 50% of cucumber circumference (total internal volume was 100 mL).
The major finding of this study is that a 50-Hz bioimpedance analyzer can detect a composition change within a biological cylinder, indicating that the electrical current passes through the entire cucumber, not only the peripheral “FFM” but also the central content, and recruits the corresponding conducting material. This is a requirement for extrapolation of our cucumber model to the human torso model. There are reports of segmental bioimpedance analysis to detect fat-free tissue.14,15 Our study explores relations between changes in the segmental ratio of FFM to volume and impedance. Both applications of regional BIA rely on the same principle. Lean tissue can be considered an electrical volume that can be compared to physical volume, with the difference related to non-conductive tissue. In a simple two-compartment model, the higher the difference in resistance between the two compartments, the greater the relative amount of current that will pass through the less resistive conductor. Because corn oil is non-conductive, virtually all of the electrical current presumably went through the conducting media of saline and cucumber mass. The distinction between lean and non-lean mass by BIA would be closer to that defined by anatomy. In the case of the human torso, for which the cucumber was used as a biological validation, the non-conductive material is comprised mainly of adipose tissue.
FIG. 4. Relation between porcine adipose tissue and surface impedance, with an electrode length of 50% of cucumber circumference.
544
He et al.
Nutrition Volume 19, Number 6, 2003
We proved our cucumber model is a valid biological model through experiment 1. Our cucumbers responded to steaming in much the same way as an earlier potato biological model validated by Lukaski using similar measurements.19 The steam-induced resistance drop in our cucumber model was 77% and reactance drop was 98%, which was close to the 76% and 100% drop documented in cooked versus raw potatoes. The results suggested that steaming enhances conductivity and abolishes reactance by destroying the cell compartments. To our knowledge, our further experiments are novel and have not been conducted previously. We found that surface impedance changed proportionally with the quantity of oil inside the cucumber. This relation was reproduced in cucumbers of different sizes with variable electrode length to circumference ratios. Therefore, the relation between impedance and internal composition is an inherent characteristic of this model, even though electrode length did affect the absolute BIA readings (data not shown). Because variation in corn oil portion in our model is equal to the variation in FFM relative to the total physical volume, a linear relation between the ratio of FFM to volume and impedance can also be interpreted as a linear relation between electrical volume and impedance regardless of total physical volume. Because fat in humans exists largely in solid form within adipose tissue, we also evaluated our model by testing solid porcine adipose tissue by methods similar to those to test liquid corn oil. We continued to see a linear relation between surface impedance and internal fat content when the fat was in a solid form regardless of its position within the BIA measurement range. These studies in cucumbers did not establish that the changes would be similar in humans. Our goal of generating a method using BIA for the clinical prediction of visceral fat content in future studies faces many potential roadblocks. Although these results showed that surface impedance reflects changes in conducting volume (electrical volume) within a segment the size of a cucumber, it remains uncertain whether the larger volume of the human torso would behave similarly. If so, it may be possible to determine non-conductive tissue (fat) volume inside an anatomic segment if one can determine the physical volume with a simple method such as anthropometry. Determining physical volume may be possible if the torso is divided into multiple cylindrical cones. This method of truncating abdominal segments is used in the current magnetic resonance imaging protocol for the whole body and was previously validated.20 Subtracting conductive volume from total physical volume would provide residual volume, which in the human abdomen is composed mainly of adipose tissue and bone. We do not expect that any method using BIA alone will be capable of further separating adipose tissue into subcutaneous and visceral compartments. However, we recently found21 a strong association between subcutaneous adipose tissue on magnetic resonance imaging and an anthropometrics model proposed by Wang et al.22 This suggests that a combination of measurements, including multiple abdominal circumferences, spacing gaps, and skinfold thicknesses, can be used to derive reliable predictions of total abdominal volume and subcutaneous adipose tissue volume. Therefore, it may be possible to estimate VAT size through subtraction. Our model supports the hypothesis that measurement of surface impedance in combination with variables of volume can be used to predict segmental internal fat content should a standard equation be derived to predict electrical volume. However, extrapolation of our simple model to measurement of the human visceral fat compartment requires further investigation, because other variables, such as irregularity of abdominal shape and separation of “subcutaneous” and “visceral” adipose tissue compartments, may affect
the result. Nevertheless, the current report is a promising first step toward the goal of using surface impedance by BIA to determine visceral fat content and its changes over time.
REFERENCES 1. Brochu M, Starling RD, Tchernof A, et al. Visceral adipose tissue is an independent correlate of glucose disposal in older obese postmenopausal women. J Clin Endocrinol Metab 2000;852:2378 2. Larsson B, Svardsudd K, Welin L, et al. Abdominal adipose tissue distribution, obesity, and risk of cardiovascular disease and death: 13 year follow up of participants in the study of men born in 1913. Br Med J 1984;288:1401 3. Nakamura T, Tokunaga K, Shimomura I, et al. Contribution of visceral fat accumulation to the development of coronary artery disease in non-obese men. Atherosclerosis 1994;107:239 4. Pascot A, Despres JP, Lemieux I, et al. Deterioration of the metabolic risk profile in women. Respective contributions of impaired glucose tolerance and visceral fat accumulation. Diabetes Care 2001;24:902 5. Pascot A, Despres JP, Lemieux I, et al. Contribution of visceral obesity to the deterioration of the metabolic risk profile in men with impaired glucose tolerance. Diabetologia 2000;43:1126 6. Thomas EL, Brynes AE, McCarthy J, et al. Preferential loss of visceral fat following aerobic exercise, measured by magnetic resonance imaging. Lipids 2000;35:769 7. Ross R, Shaw KD, Rissanen J, et al. Sex differences in lean and adipose tissue distribution by magnetic resonance imaging: anthropometric relationships. Am J Clin Nutr 1994;59:1277 8. Clasey JL, Bouchard C, Teates CD, et al. The use of anthropometric and dual-energy X-ray absorptiometry (DXA) measures to estimate total abdominal and abdominal visceral fat in men and women. Obes Res 1999;7:256 9. Kamel EG, McNeill G, Van Wijk MC. Usefulness of anthropometry and DXA in predicting intra-abdominal fat in obese men and women. Obes Res 2000;8:36 10. Ross R, Rissanen J, Pedwell H, Clifford J, Shragge P. Influence of diet and exercise on skeletal muscle and visceral adipose tissue in men. J Appl Physiol 1996;81:2445 11. Fowler PA, Fulle MF, Glasbey CA, et al. Total and subcutaneous adipose tissue in women: the measurement of distribution and accurate prediction of quantity by using magnetic resonance imaging. Am J Clin Nutr 1995;54:18 12. Baumgartner RN, Rhyne RL, Troup C, Wayne S, Garry PJ. Appendicular skeletal muscle areas assessed by magnetic resonance imaging in older persons. J Gerontol 1992;47:M67 13. Engelson ES, Kotler DP, Tan Y, et al. Fat distribution in HIV-infected patients reporting truncal enlargement quantified by whole-body magnetic resonance imaging. Am J Clin Nutr 1999;69:1162 14. Heymsfield SB, Gallagher D, Grammes J, et al. Upper extremity skeletal muscle mass: potential of measurement with single frequency bioimpedance analysis. Appl Radiat Isot 19989;49:473 15. Lukaski HC. Assessing regional muscle mass with segmental measurements of bioelectrical impedance in obese women during weight loss. Ann NY Acad Sci 2000;904:154 16. Kotler DP, Burastero S, Wang J, Pierson RN, Jr. Prediction of body cell mass, fat-free mass, and total body water with bioelectrical impedance analysis: effects of race, sex, and disease. Am J Clin Nut 1994;64:489S 17. Bracco D, Thiebaud D, Chiolero RL, et al. Segmental body composition assessed by bioelectrical impedance analysis and DEXA in humans. J Appl Physiol 1996;81:2580 18. Fuller NJ, Fewtrell MS, Dewit O, Elia M, Wells JC. Segmental bioelectrical impedance analysis in children aged 8 –12 y: 2. The assessment of regional body composition and muscle mass. Int J Obes Relat Metab Disord 2002;26:692 19. Lukaski HC. Biological indexes considered in the derivation of the bioelectrical impedance analysis. Am J Clin Nutr 1996;64:397S 20. Abate N, Burns D, Peshock RM, Garg A, Grundy SM. Estimation of adipose tissue mass by magnetic resonance imaging: validation against dissection in human cadavers. J Lipid Res 1994;35:1490 21. He Q, Wang J, Engelson ES, Kotler DP. Ability of anthropometrics model to track changes in subcutaneous adipose tissue area (abstract). FASEB J 2002;16: A1024 22. Wang J, Thornton JC, Russell M, et al. Asians have lower body mass index (BMI) but higher percent body fat than do whites: comparisons of anthropometric measurements. Am J Clin Nutr 1994;60:23
(For an additional perspective, see Editorial Opinions.)