Using The Bioimpedance Spectroscopy For Diagnosis Of Malnutrition in Chronic Kidney Disease Stage 5—Is It Useful

Abstract

1. Objective

Malnutrition is common in chronic kidney disease stage 5 (CKD5) and has negative clinical impacts. The present study aims to evaluate bioimpedance spectroscopy (BIS) in diagnosing malnutrition in CKD5 including hemodialysis and peritoneal dialysis patients (CKD5D) using cutoff values for fat-free mass index (FFMI) according to the Global Leadership Initiative on Malnutrition criteria. Dual-energy X-ray absorptiometry (DXA) was used as a reference method.

Design and Methods

We performed a single-center cross-sectional diagnostic study of 90 patients with CKD5 or CKD5D.

2. Results

BIS-derived FFMI estimates were significantly higher compared with those obtained by DXA (18.5 6 2.6 vs.17.8 6 2.0, P, .05). The mean difference in FFMI estimates between the methods (DXA–BIS) and Bland-Altman 95% limits of agreements is –0.38 (2.76, –3.52) kg/m2. Overhydration (B 5 0.67, P, .001), age (B 5 0.02, P 5 .037), and interactions between overhydration and CKD5 subgroups (P 5 .034) independently predicted bias in BIS-derived FFMI. BIS-derived FFMI showed poor sensitivity (64%) and positive predictive value (48%) in diagnosing malnutrition in the present study population.

3. Conclusion

The present study showed a limited agreement between estimates of FFMI derived by BIS and DXA due to a large interindividual variation. Using BIS as a clinical tool for assessing FFMI has limited accuracy and poor sensitivity in diagnosing malnutrition in patients with CKD5 and CKD5D.

Click here to know what is the Cistanche

Introduction

MALNUTRITION IS COMMON in patients with chronic kidney disease (CKD) and has negative clinical impacts comprising increased hospital stay duration, incidence of infections, morbidity, and mortality.1-3 The International Society of Renal Nutrition and Metabolism has developed a framework for nutritional status assessment and diagnosing malnutrition in CKD, namely protein-energy wasting (PEW).4 Diagnosis of PEW includes four categories covering body weight, muscle mass, serum chemistry, and dietary intake. At least three out of these four categories must be met for the diagnosis of PEW. The prevalence of PEW according to subjective global assessment or malnutrition inflammation score in CKD stage 5 (CKD5) including patients undergoing maintenance dialysis (CKD5D) varies between 11% and 54%.3

Recently, the Global Leadership Initiative on Malnutrition (GLIM) has published criteria for the diagnosis of malnutrition based on phenotypic criteria (weight loss, low body mass index, and reduced muscle mass) and etiologic criteria (reduced food intake or assimilation and inflammation).5 For the diagnosis of malnutrition, at least one phenotypic and one etiologic criterion is required. Mild to moderate inflammation is common in patients with CKD5. Accordingly, one phenotypic criterion as reduced muscle mass can be used to diagnose malnutrition in these patients. Both these diagnostic frameworks, PEW and GLIM include measurements of body composition (BC). Assessment of BC has received increasing attention owing to the negative impact of the loss of fat-free mass (FFM), muscle mass, and strength on patient survival.6

Assessment of BC can be done using physical examinations and anthropometric measurements of circumferences and skinfold thickness or by technical methods such as dual-energy X-ray absorptiometry (DXA) and bioelectrical impedance analysis (BIA). Although DXA is the most commonly used reference method,7 the high cost and reduced availability make it less appropriate as a screening or bedside method in a daily clinical setting. Anthropometry, on the other hand, is easy to perform and cheap, but has low validity.8 Using multifrequency bioimpedance spectroscopy (BIS) is a simple, reproducible, and relatively inexpensive BIA method that can easily be used in both outpatient settings and bedside in hospitalized patients.9,10 Owing to the major impact of malnutrition on morbidity and mortality in CKD patients, the use of BIS for BC assessment has earned interest as a potentially valuable tool for diagnosis of malnutrition in CKD patients.9 However, the quality of existing evidence regarding the validity of using BIS in CKD patients is low and further studies are recommended by the American Society for Parenteral and Enteral Nutrition.11

The present study aims to evaluate the use of BIS for BC assessment and diagnosis of malnutrition in CKD5 patients using cutoff values for fat-free mass index (FFMI) according to GLIM criteria with DXA as a reference method.

Herba Cistanche

Materials and Methods

1. Study Participants

In the present single-center cross-sectional study, 93 patients were recruited consecutively between April 2015 and June 2018 (during scheduled visits between April 22 and June 17, 2015, August 31 and November 30, 2015, and January 20 and February 29, 2016) from our nephrology outpatient clinic, the peritoneal dialysis (PD) and hemodialysis (HD) clinic. The regional ethics committee as well as the radiation protection committee approved the study protocol. The research was conducted by the Helsinki Declaration. Eligible study subjects were informed about the study in both oral and written form. Consent was provided by all the subjects who participated in the study after a full and comprehensive explanation of the study.

Inclusion criteria were adults (.18 years) with CKD5 or CKD5D followed regularly.

Exclusion criteria were inability to independently take part in the BC measurements (unable to communicate in Swedish, dementia or physical impairment), amputation, pacemaker, pregnancy, and body mass index (BMI),16 or .34 kg/m2.

2. Measurements

Height was measured barefoot to the nearest 0.5 cm using a wall-mounted measure. Body weight was measured in underwear to the nearest 100 g using a calibrated Tanita MC-180 MA.

3. Dual-Energy X-ray Absorptiometry

Whole-body DXA measurements were performed using a Lunar Prodigy scanner (GE Healthcare, Danderyd, Sweden, enCore software version 14.1) where BC is determined by using the different attenuation (or absorption) characteristics of bone, fat, and fat-free soft tissue mass. When low-dose X-rays of two different energies pass through the body, an image is created as the photon detector measures the different attenuation of the low and high X-ray energy by the soft tissue and bone. DXA is considered as a reference method in BC measurement owing to its high degree of precision, with a coefficient of variation (CV) of 1%-2% CV for lean soft tissue.7

Cistanche supplement

4. Multifrequency Bioelectrical Impedance Analysis

All patients in the study were measured in a standardized manner according to the device instruction manual of the Body Composition Monitor with software version 3.2.x (Fresenius Medical Care, Bad Homburg, Germany) which is a whole-body BIS at 50 different frequencies between 5 kHz and 1 MHz. Volume overload is determined in liters based on a three-compartment physiological tissue model which differentiates between normohydrated lean tissue mass (LTM), adipose tissue mass, and a virtual overhydration (OH) compartment. This model assumes fixed hydration of LTM and adipose tissue mass for the calculation of a ‘‘normohydration weight’’. In each patient, DXA and BIS measurements were performed concurrently (within 10 minutes). Patients were asked to not eat a big meal, drink or perform any rigorous physical activity before measurements, to lie down for .10 minutes before measurement, and remove jewelry and glasses. In PD patients, peritoneal fluid was drained from the abdomen before measurements. In HD patients, measurements were done before attending the midweek dialysis session. Measurements were done on the nonfistula or noncentral dialysis access side. System tests according to device instructions were made regularly.

5. Body Composition and Assessment of Fat-Free Mass

As FFM is not obtainable in the output data of either DXA or BIS-derived measurements, FFM was calculated as follows:

6. Statistics

Power calculations were made assuming the study participants have a mean FFM of around 55 kg with a standard deviation of 6 10 kg and a mean height of 1.70 m based on previous publications.12,13 Given an a of 5% and statistical power (1 2 b) of 80%, the minimum number of subjects to be included in the study to detect a difference of 3 kg in FFM (z1 kg/m2 FFMI) was 82. Presented values are means 6 standard deviations unless stated otherwise. Statistical significance was set at the level of P, .05. Student’s paired t-tests were used for comparing differences in continuous data. Correlations between continuous data were calculated using Pearson’s or Spearman’s test when appropriate. Bias in BIS was calculated by subtracting the DXA-derived FFMI from those derived by BIS (delta FFMI 5 DXA–BIS). The Bland-Altman method was used to assess the interindividual variability between the DXA and BIS-derived FFMI estimates.14 Multiple linear regression analysis was performed to assess factors independently predicting bias in BIS measurements of FFMI. Statistical analyses were performed using the SPSS Statistics Data Editor (IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp.).

Cistanche tubulosa

Discussion

The main findings of the present study were that BISderived BC estimates overestimated FFMI compared to those obtained by DXA. The accuracy of BIS-derived FFMI was moderate. BIS-derived FFMI showed poor sensitivity and positive predictive value in diagnosing malnutrition in the present study population. Parameters of BC derived by BIS showed statistically significant lower FM and higher FFM and FFMI levels compared to those derived by DXA despite low mean differences between the methods for these parameters. This can be explained by large interindividual variations in BC estimates derived by BIS.

Using DXA as the reference test, Bross et al. showed that the single frequency BIA- Kushner equation underestimated FM despite being the most accurate of the BIA equations (Kushner, Lukaski, and Segal equations) in estimating FM in 118 maintenance HD patients.15 Although BIS overestimated FM in our cohort, there was no significant difference in BIS-derived FM estimates to those derived by DXA in the HD subgroup. However, as BIA devices provide estimates of BC by using different equations and algorithms that vary based on the manufacture of the device and the BIA technique (single frequency BIA, multifrequency BIA, or BIS), comparing results of studies conducted with different BIA devices is dissuaded and should be interpreted with great caution.

To our knowledge, few studies have evaluated the use of BIS and multifrequency BIA for the estimation of BC in patients with CKD5 and CKD5D using DXA as a reference test. F€urstenberg and Davenport compared multifrequency BIA-derived FFM estimates to those obtained by DXA in 104 patients on PD and 53 on HD.16,17 Although the mean difference between the two methods was markedly lower (–0.16 and 1 0.53 kg in PD and HD patients, respectively) compared to the present study, the interindividual variations were large (95% limits of agreement –6.96 to 6.64 kg and –8.10 to 9.11 kg in PD and HD patients, respectively) and comparable to those in the present study. Similar to the findings in the present study, Zhou et al. showed that BIS-derived BC estimates overestimated FFM compared to those obtained by DXA in 120 nondialysis patients with CKD.13 The agreement between the two methods was limited with a substantial mean difference (–2.8 kg) and interindividual variation (95% limits of agreement –12 to 6.5 kg). Likewise, Vine et al. showed limited agreement between BIS-derived FFM estimates to those obtained by DXA in 12 patients with CKD5 and 16 with CKD5D on HD.18 The mean differences between the methods were –1.9 and 0.9 kg in CKD5 and CKD5D groups, respectively, with large interindividual variations (95% limits of agreement –14.7 to 10.9 kg and –9.7 to 11.6 kg in CKD5 and CKD5D, respectively). Also, Popovic et al. showed wide interindividual variations (95% limits of agreement –11.8 to 10.8 kg) despite a low mean difference (–0.3 kg) comparing BIS-derived FFM estimates to those derived by DXA in 72 patients with CKD5D on PD.12 Rymarz et al. showed that BIS overestimated FM in comparison with DXA. This finding is not in accord with our finding as we showed no significant difference in BISderived FM estimates to those derived by DXA in the HD subgroup. 19 This discrepancy between the two studies could, at least in part, be explained by differences in hydration status as, unlike the present study, measurements with DXA and BIS were performed at different time points. However, our data regarding FM in the HD subgroup are comparable to those reported in previous studies. Unfortunately, FFM data were not reported by Rymarz et al. to perform further comparisons with the present study.

Collectively, our results comparing BIS-derived FFMI estimates to those derived by DXA in patients with CKD5 and CKD5D are in good agreement with the earlier findings showing large interindividual variation that reduces the accuracy of using BIS as a clinical tool for assessing FFM in CKD5 patients.

Cistanche capsules

Few studies have investigated factors associated with the bias in BIS-derived FFM estimates in CKD5 and CKD5D.12,13,18 These factors varied among the studies. In line with Zhou et al. and Popovic et al., OH in the present study showed a significant independent positive correlation to the bias in BIS-derived estimates of FFMI.12,13 The degree of OH in the present study was moderate and comparable to those reported in the previous studies. Popovic et al. showed that OH was disproportionally correlated to FM obtained by DXA and suggested that OH might have caused an error in the DXA analysis. It has been suggested that hydration status might affect DXA accuracy owing to the programmed assumption of a constant FFM hydration.20 Large changes in hydration status may alter the attenuation of FFM resulting in its overestimation and thereby underestimation of FM.20 In line with this, the degree of OH in the present study showed an inverse significant correlation to FM estimate by DXA (r 5 –0.25, P 5 .02). Still, DXA-derived BC measurements showed lower levels of FFM and higher FM compared to those derived by BIS suggesting a possible impact of other factors on BC estimates by DXA and BIS. Furthermore, in the present study, we found a significant unique independent interaction effect of OH and subgroups category that reduced the bias in BIS-derived estimates of FFMI in CKD5 and PD subgroup compared to HD as the reference category for subgroups. We found no correlation between OH and FFMI derived by BIS or DXA. Thus, this novel finding needs to be investigated in further studies.

In the present study, age showed an independent small positive significant association with the bias in BIS-derived FFMI. Vine et al. showed that age and BMI were significantly correlated with the bias in BIS-derived FFM.18 However, regression analysis showed no significant association of age to BIS bias. Nevertheless, the interactions between gender and age and gender and BMI significantly predicted the bias in BIS-derived FFM. We found no such interaction effect even when we included BMI in the regression analysis in the present study (data not shown). The reason for this disparity is unclear and may be multifactorial including differences in study populations. The low proportion of females in the present study might have masked the possible impact of gender on FFMI estimates. In addition, the findings of the regression analysis by Vine et al. should be interpreted with caution due to the small study population.18

As anticipated, the diagnostic accuracy of BIS for malnutrition according to GLIM criteria was poor owing to its poor sensitivity and major misclassification of nutritional status by BIS-derived FFMI. The poor sensitivity of BIS can be explained, at least in part, by the overestimation of FFMI by BIS. However, the proportion of patients categorized as having malnutrition by BIS was larger compared to DXA. This can be explained by the large interindividual variation in estimates of FFMI derived by BIS. In addition, one can speculate that the reference method, DXA, might not be optimal in this group of patients due to the possible impact of different factors including OH on its accuracy.

The present study has some strengths: first, in contrast to some studies,16,17,19 measurements with DXA and BIS were done concurrently to avoid any changes in body weight and composition biasing the comparability of the two methods. Second, all measurements of BC were done in a standardized manner by two technicians throughout the study which minimize bias caused by interobserver variability. In addition, given the heterogeneity of the CKD5 group including predialysis patients and patients on maintenance dialysis with different modalities (HD and PD), we were able to evaluate the use of BIS in a cohort comprising the whole spectrum of CKD5 in a single-center study unlike to the previous studies.12,13,15-19,21 This allows for more reliable comparison and identifying major differences between the subgroups. Furthermore, in the present study, we extended our analyses and evaluated the diagnostic accuracy of BIS for malnutrition according to GLIM criteria. However, the present study has some limitations. Owing to practical challenges in the clinical setting, we performed measurements of BC before the start of Hn contrary to the recommendations of the device (at least 30 minutes after a completed dialysis session) and several studies.17,18,22 However, despite performing BC measurements before the start of HD, the degree of OH in subgroup HD was moderate and comparable to subgroups CKD5 and PD. In addition, the bias in BIS-derived FFMI was the least in subgroup HD. Still, one can speculate that performing BC measurements after a completed dialysis session would have improved the accuracy in BISderived FFMI in subgroup HD.

The Cistanche benefits

In conclusion, the present study showed a limited agreement between estimates of FFMI derived by BIS and DXA owing to a large interindividual variation. Using BIS as a clinical tool to assess FFMI is associated with low accuracy and poor sensitivity in diagnosing malnutrition in patients with CKD5 and CKD5D. The degree of OH and the interaction of OH with the subgroups category showed an independent association with the bias in BIS-derived FFMI, yet their impacts were moderate. Further studies are needed to explore probable additional factors that may impact the agreement between the two methods.

References

1. Carrero JJ, Stenvinkel P, Cuppari L, et al. Etiology of the protein-energy wasting syndrome in chronic kidney disease: a consensus statement from the International Society of Renal Nutrition and Metabolism (ISRNM). J Ren Nutr. 2013;23:77-90.

2. Gracia-Iguacel C, Gonz'alez-Parra E, P'erez-Gomez MV, et al. Prevalence of protein-energy wasting syndrome and its association with mortality in hemodialysis patients in a center in Spain. Nefrologia. 2013;33:495-505.

3. Carrero JJ, Thomas F, Nagy K, et al. The global prevalence of protein energy wasting in kidney disease: a Meta-analysis of Contemporary Observational studies from the International Society of renal nutrition and Metabolism. J Ren Nutr. 2018;28:380-392.

4. Fouque D, Kalantar-Zadeh K, Kopple J, et al. A proposed nomenclature and diagnostic criteria for protein-energy wasting in acute and chronic kidney disease. Kidney Int. 2008;73:391-398.

5. Cederholm T, Jensen GL, Correia M, et al. GLIM criteria for the diagnosis of malnutrition - a consensus report from the global clinical nutrition community. Clin Nutr (Edinburgh, Scotland). 2019;38:1-9.

6. Cruz-Jentoft AJ, Bahat G, Bauer J, et al. Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing. 2019;48:16-31.

7. Earthman CP. Body composition tools for assessment of adult malnutrition at the bedside: a Tutorial on research Considerations and clinical applications. JPEN J Parenter Enteral Nutr. 2015;39:787-822.

8. Prado CMM, Heymsfield SB. Lean tissue imaging. J Parenter Enteral Nutr. 2014;38:940-953.

9. L'opez-G'omez JM. Evolution and applications of bioimpedance in managing chronic kidney disease. Nefrologıa (English Edition). 2011;31:630-634.

10. Lukaski HC. Evolution of bioimpedance: a circuitous journey from estimation of physiological function to assessment of body composition and a return to clinical research. Eur J Clin Nutr. 2013;67:S2-S9.

11. Sheean P, Gonzalez MC, Prado CM, McKeever L, Hall AM, Braunschweig CA. American Society for Parenteral and Enteral nutrition clinical Guidelines: the validity of body composition assessment in clinical populations. JPEN J Parenter enteral Nutr. 2020;44:12-43.

12. Popovic V, Zerahn B, Heaf JG. Comparison of dual-energy X-ray absorptiometry and bioimpedance in assessing body composition and nutrition in peritoneal dialysis patients. J Ren Nutr. 2017;27:355-363.

13. Zhou Y, H€oglund P, Clyne N. Comparison of DEXA and bioimpedance for body composition measurements in Nondialysis patients with CKD. J Ren Nutr. 2019;29:33-38.

14. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet (London, England). 1986;1:307-310.

15. Bross R, Chandramohan G, Kovesdy CP, et al. Comparing body composition assessment tests in long-term hemodialysis patients. Am J Kidney Dis. 2010;55:885-896.

16. F€urstenberg A, Davenport A. Assessment of body composition in peritoneal dialysis patients using bioelectrical impedance and dual-energy x-ray absorptiometry. Am J Nephrol. 2011;33:150-156.

17. F€urstenberg A, Davenport A. Comparison of multifrequency bioelectrical impedance analysis and dual-energy X-ray absorptiometry assessments in outpatient hemodialysis patients. Am J Kidney Dis. 2011;57:123-129.

18. Vine SM, Painter PL, Kuskowski MA, Earthman CP. Bioimpedance spectroscopy for the estimation of fat-free mass in end-stage renal disease. E SPEN Eur E J Clin Nutr Metab. 2011;6:e1-e6.

19. Rymarz A, Szamotulska K, Niemczyk S. Comparison of skinfold Thicknesses and bioimpedance spectroscopy to dual-energy X-ray absorptiometry for the body fat measurement in patients with chronic kidney disease. Nutr Clin Pract. 2017;32:533-538.

20. Prado CM, Heymsfield SB. Lean tissue imaging: a new era for nutritional assessment and intervention. JPEN J Parenter enteral Nutr. 2014;38:940-953.

21. Molfino A, Don BR, Kaysen GA. Comparison of bioimpedance and dual-energy x-ray absorptiometry for measurement of fat mass in hemodialysis patients. Nephron Clin Pract. 2012;122:127-133.

22. Panorchan K, Nongnuch A, El-Kateb S, Goodlad C, Davenport A. Changes in muscle and fat mass with hemodialysis detected by multifrequency bioelectrical impedance analysis. Eur J Clin Nutr. 2015;69: 1109-1112.

Sintra Eyre, RD,Ingvar Bosaeus, MD, PhD, Gert Jensen, MD, PhD, and Aso Saeed, MD, PhD